ArticlePDF Available

Differences between girls and boys in emerging language skills: Evidence from 10 language communities

Authors:
  • University Medical Center Goettingen / Karolinska Institutet / Medical University of Graz

Abstract and Figures

The present study explored gender differences in emerging language skills in 13,783 European children from 10 non-English language communities. It was based on a synthesis of published data assessed with adapted versions of the MacArthur-Bates Communicative Development Inventories (CDIs) from age 0.08 to 2.06. The results showed that girls are slightly ahead of boys in early communicative gestures, in productive vocabulary, and in combining words. The difference increased with age. Boys were not found to be more variable than girls. Despite extensive variation in language skills between language communities, the difference between girls and boys remained. This suggests that the difference is caused by robust factors that do not change between language communities.
Content may be subject to copyright.
326
British Journal of Developmental Psychology (2012), 30, 326–343
C2011 The British Psychological Society
The
British
Psychological
Society
www.wileyonlinelibrary.com
Differences between girls and boys in emerging
language skills: Evidence from 10 language
communities
M˚
arten Eriksson1, Peter B. Marschik2, Tiia Tulviste3,
Margareta Almgren4,MiguelP
´
erez Pereira5, Sonja Wehberg6,
Ljubica Marjanoviˇ
c-Umek7, Frederique Gayraud8,
Melita Kovacevic9and Carlos Gallego10
1Faculty of Health and Occupational Studies, University of G¨
avle, Sweden
2Institute of Physiology, Center for Physiological Medicine, Medical University of
Graz, Austria
3Institute of Psychology, University of Tartu, Estonia
4University of the Basque Country, Vitoria-Gasteiz, Spain
5Department of Developmental and Educational Psychology, University of Santiago
de Compostela, Spain
6Institute of Language and Communication, University of Southern Denmark,
Denmark
7Department of Psychology, University of Ljubljana, Slovenia
8Dynamics of Language, University of Lyon, France
9Laboratory for Psycholinguistic Research, University of Zagreb, Croatia
10Department of Psychology, Complutense University of Madrid, Spain
The present study explored gender differences in emerging language skills in 13,783
European children from 10 non-English language communities. It was based on a
synthesis of published data assessed with adapted versions of the MacArthur-Bates
Communicative Development Inventories (CDIs) from age 0.08 to 2.06. The results
showed that girls are slightly ahead of boys in early communicative gestures, in productive
vocabulary, and in combining words. The difference increased with age. Boys were not
found to be more variable than girls. Despite extensive variation in language skills
between language communities, the difference between girls and boys remained. This
suggests that the difference is caused by robust factors that do not change between
language communities.
Correspondence should be addressed to Associate professor M˚
arten Eriksson, Faculty of Health and Occupational Studies,
University of G¨
avle, SE-801 76 G¨
avle, Sweden (e-mail: marten.eriksson@hig.se).
DOI:10.1111/j.2044-835X.2011.02042.x
Differences between girls and boys in early language 327
Girls and boys are similar at birth in that neither gender talk. In their well-known review
of gender differences in cognitive functioning, Maccoby and Jacklin (1974) concluded
that girls matured more rapidly in verbal abilities than boys but did not find solid evidence
for this divergence before the age of 11 years. Hyde and Linn (1988) reported a small girl
advantage but no developmental trend in a meta-analysis of gender differences in verbal
ability including the studies of Maccoby and Jacklin (1974) as well as later ones. No child
younger than 3 years was included in these studies and participants were restricted to
those living in the United States or Canada. In a more recent study Bornstein, Hahn,
and Haynes (2004) reported that American girls scored higher than American boys on a
multitude of language measures, including spontaneous speech, caregiver reports, and
formal testing from age 2 to 5 years. Thus, although many studies indicate that girls are
slightly ahead of boys in verbal skills, there seems to be some confusion about the timing
when girls start to outperform boys. Further, the majority of research on early child
language acquisition has been carried out in a mono-language, mono-cultural setting
(English/American). Such studies are particularly vulnerable to confounding variables in
questions of nature versus nurture. Therefore, comparative studies of child language in
multiple linguistic and cultural settings are considered to be of special value (Bornstein,
2002; Bornstein et al., 2004). The present study extends previous research on gender
differences in language by investigating emerging language skills in a large sample of
children from 10 non-English language communities, thereby making it less vulnerable
to cultural and linguistic biases. Both differences in mean language skills and differences
in variability between girls and boys will be studied.
Gender differences in language for children before age 3
Comprehensive studies on gender differences in verbal ability for children below the
age of 3 years have mainly been performed with parent report instruments, such as
the MacArthur-Bates Communicative Development Inventories (CDIs) (e.g., Fenson
et al., 1994, 2007). Based on the CDI norming study, Fenson and colleagues (1994,
2007) reported a girl advantage that accounted for about 1–2% of the variance in a study
that included 2,550 American children aged 0.08–2.06 (year; months). The girl advantage
concerned gestures as well as vocabulary comprehension and vocabulary production.
However, an interaction between gender and age (the difference increasing with age)
was only reported for vocabulary production and only in children aged 0.08–1.04. The
scarcity of reported interactions between gender and age on language skills is surprising
because gender differences are recurrently reported and an interaction with age is to be
expected for any emerging skill in which girls and boys are going to differ. In another
large-scale study based on a short version of the CDI, Galsworthy, Dionne, Dale, and
Plomin (2000) found a girl advantage in a study of over 3,000 2-year-old British children
that accounted for about 3% of the variance in vocabulary production. Thus, meta-
analyses and large-scale parental reports suggest that there is a consistent girl advantage
in early language acquisition, at least for children acquiring English. This advantage has
only been reported occasionally to interact with age.
The CDI instrument has been adapted and normed for several non-English cultures
and languages. The results from these norming studies are complex concerning differ-
ences between girls and boys. Many reported a girl advantage for at least one language
skill, primarily word production (for Danish, Bleses et al. 2008; for Estonian, Tulviste,
2007; for French, Kern, 2007; for Slovene, Marjanoviˇ
c-Umek, Fekonja, Kranjc, & Bajc,
2008; for Spanish, Gallego & Mariscal, 2008), more rarely for word comprehension
(for Basque, Garc´
ıa, Ezeizabarrena, Almgren, & Errarte, 2005; for Danish, Bleses et al.,
328 M˚
arten Eriksson et al.
2008) and gestures (for Danish, Bleses et al., 2008; for Swedish, Eriksson & Berglund,
1999). In contrast, no gender difference was reported by Jackson-Maldonado and
colleagues among children acquiring Mexican-Spanish (Jackson-Maldonado, Thal, Bates,
Marchman, & Gutierrez-Clellen, 1993), by Berglund and Eriksson (2000) for Swedish-
speaking children or by Kern & Gayraud (2007) for preterm children acquiring French.
However, some of these studies are limited by small sample size and the inclusion of
age is inconsistent (different age intervals are used, if age is included at all). Because age
explains a major part of the variance in children’s language skills, it is crucial to include
age in a consistent way in statistical models.
Few cross-linguistic and cross-cultural studies on early language acquisition have been
performed in relation to gender differences. In one such rare study, Bornstein and Cote
(2005) reported that girls aged 1.08 had larger vocabularies than boys of the same age
across languages as well as urban and rural settings. The study was conducted in three
countries comparing children from the United States, Argentina, and Italy.
Social explanations to gender differences in language
Many theories of gender differences (e.g., the gender role socialisation theory by Jacklin
& Baker, 1993 or the social role theory by Eagly, Wood, & Diekman, 2000) stress the
influence of the social environment on various cognitive domains, including language.
For example, parents typically choose gender-typed toys even for very young children
(Fagot, 1995). Caldera, Huston, and O’Brian (1989) argued that the action-oriented
toys, more likely bought for boys, do not stimulate language in the same way as the
more caring-oriented toys normally bought for girls. In response to action play, parents
mainly produced animated sounds and verbal corrections as compared to caring play in
which parents produced more verbal interactions in forms of comments and questions.
Similarly, Bornstein and colleagues, (Bornstein, Haynes, Pascual, Painter, & Galperin,
1999; Suizzo & Bornstein, 2006) found that mothers of boys in Argentina, France, and
the United States engaged in more exploratory play than did mothers of girls, and
mothers of girls engaged in more symbolic play than did mothers of boys. Corresponding
differences were also found in the amount of exploratory and symbolic play girls and
boys themselves engaged in. This is important because several studies have shown that
amount of symbolic play in children is highly related to more advanced language (e.g.,
Bates, Benigni, Bretherton, Camaioni, & Volterra, 1979; Lyytinen, Poikkeus, & Laakso,
1997). Hence, there is a multitude of social factors that may contribute to different
acquisition of language among girls and boys. Because the particular mix of these
factors is expected to vary among different language communities (Best & Williams,
1997), gender differences in early language skills are consequently also expected to vary
among language communities if caused by the social environment. (The term language
community is used throughout this paper to refer to the specific linguistic forms and
structures constituting a specific language as well as language as an expression of
culture. Language is the prime medium for a culture to express itself. The independence
that languages can have from national borders is also acknowledged by the term.)
One purpose of the present study is to test the robustness of the gender effect in
language by comparing early language skills between girls and boys in different language
communities.
The greater male variability hypothesis
Since the days of Ellis (Ellis, 1894), it has been argued at times that men are more
variable than women in cognitive functioning, which is often referred to as the greater
Differences between girls and boys in early language 329
male variability hypothesis (i.e., greater variability among males in intelligence). The
evidence for this hypothesis varies. Arden and Plomin (2006) found support for greater
variance of intelligence among boys at age 3 to 10 but not at age 2. Homogeneity of
variance between genders has usually been found for most verbal tests (see Feingold,
1992 for a review). Thus, inconsistent support for the greater male variability hypothesis
could either depend on the investigated age period, on the cognitive domain, or both.
Knowledge of equal or unequal variance between genders might have important applied
consequences in determining cut-offs in screening programs. For example, boys are often
over-represented in the lower percentiles when screening for language delay (Law, Boyle,
Harris, Harkness, & Nye, 1998). This observation might reflect a lower mean performance
in the measured language skills among boys in comparison with girls, a greater variability
among boys, or both. Whereas equal means and equal variance between genders would
call for unisex norms, findings of equal means and heterogeneous variance, or different
means (with or without a difference in variance) would be a more complicated matter.
It might indicate that one gender is more vulnerable and, therefore, more in need for
special interventions and unisex norms should be kept, or that unisex norms would
result in many false positives within the gender with many individuals in the lower tail
(boys) and separate norms should be used instead. Long-term follow-ups of girls and
boys at the lower tail of each gender would be needed to delineate these alternatives if
heterogeneous variance or homogeneous variance and a difference in means are found.
Hence, the male variability hypothesis is not only of mere historical interest but also
has current relevance for clinical practice. Therefore, another purpose of the present
study is to test the hypothesis that boys are more variable than girls in language skills at
language onset.
The present study
This study extends earlier research on gender differences by merging data from
previously published studies on early language skills in girls and boys from 10 non-
English European language communities. The inclusion of both language community
and age as independent variables in the analyses and in interaction with gender allows
us to control for these variables in a coherent way. Variety of differences between girls
and boys across the 10 language communities (expressed as an interaction between
gender and language community) would be evidence of causes within the language
communities, for example in terms of linguistic variables or social variables such as
different ways to raise and talk to girls and boys. Neurodevelopmental factors such as
earlier brain lateralization in girls (Crow, 1998) do certainly not differ across language
communities and could not account for such a result.
In summary, this study tests the following four hypotheses: (1) girls are ahead of
boys in emerging language skills; (2) the gender difference in language interacts with
language community; (3) the gender difference in language interacts with age group;
and (4) young boys are more variable in language skills than young girls.
Method
This study is a synthesis of normed data from previously published studies (see Language
Communities for references). All participants were assessed with adapted versions of
the MacArthur-Bates CDI.
330 M˚
arten Eriksson et al.
Ta b l e 1 . Number of girls and boys over two CDI forms and 10 language communities (N=13,783)
Language community Words and gestures Words and sentences
Girls Boys Girls Boys
Austrian German 72 42 266 147
Basque 225 217 494 481
Croatian 125 125 179 197
Danish 848 864 1477 1386
Estonian 182 190 325 314
French 275 273 346 317
Galician 190 186 348 352
Slovene 443 484
Spanish 223 189 296 297
Swedish 241 224 473 420
Total 2381 2310 4697 4395
Participants
In total, 13,783 children from 10 European language communities contributed data
to this study. Of these 13,783 children 4,691 were aged 0.08–1.04 and assessed with
adapted versions of the Words and Gestures form (W&G) of the CDI. The remaining
9,092 children were aged 1.04–2.06 and assessed with adapted versions of the Words
and Sentences form (W&S) of the CDI. The proportion of girls in each age group was
51%. The proportion of girls in the different samples varied at most 3% from the total
proportion with the exception of Austrian girls who were considerably over-represented,
constituting 63% and 64% for W&G and W&S, respectively (Table 1). Birth order was
evenly distributed over girls and boys. In the infant sample (assessed with W&G), 48.3%
of the girls and 48.1% of the boys were firstborns. For the toddler sample (assessed with
W&S), 53.2% of the girls and 52.8% of the boys were firstborns. Among both infants and
toddlers, 11.8% of the girls were bilingual whereas 10.9% of the infant boys and 12.2%
of the toddler boys were bilingual.
Measures and procedure
Word comprehension and production in infants were assessed with adapted versions of
the W&G form and word production in toddlers was assessed using adapted versions
of the W&S form. These forms have a checklist format enabling a parent to mark the
particular words the child presently understands or produces. The words are grouped
by semantic category to facilitate retrieval. In addition, a similar checklist in the adapted
W&G forms assessing children’s first communicative gestures was used. Parents’ answer
(yes/no) to the question of whether the child combined words was assessed by the
adapted W&S forms. The number of items varied in the versions adapted to different
language communities. The number of gestures varied between 11 and 13 words, the
vocabulary checklist in the adapted W&G form varied between 303 (Spanish) and 688
(Austrian German) words, and in the adapted W&S between 588 and 725 words. A
major cause of the difference in length in the adapted W&G forms was that the Austrian
infant (W&G) and toddler (W&S) forms consisted of almost identical number of items.
Minor causes of differences in length were due to linguistic and cultural characteristics.
For example, the number of pronouns and kinship terms differs markedly between
Differences between girls and boys in early language 331
languages and the relevance of words denoting food items and clothing differs between
communities.
Adapted versions of the CDI were distributed to parents with children of the relevant
ages. The main method of contact across language communities was in cooperation with
educational and paediatric institutions. Some parents were contacted through random
selection from population registers (e.g., Denmark and Sweden). Most of the parents
returned the instrument by mail. Girls and boys were recruited in the same way within
each language community.
Language communities
The sample consists of 10 European language communities covering the following
languages: three Germanic languages: Austrian-German, (Marschik, Einspieler, Garzarolli,
& Prechtl, 2007), Danish (Bleses, Vach, Wehberg, Faber, & Madsen, 2007), and Swedish
(Berglund & Eriksson, 2000; Eriksson & Berglund, 1999); three Romance languages:
French, (Kern, 2007), Galician (P´
erez-Pereira, 2008; P´
erez-Pereira & Garc´
ıa-Soto, 2003;
P´
erez-Pereira & Resches, 2011), and Spanish (L´
opez Ornat et al., 2005; Mariscal, L´
opez
Ornat, Gallego, Karousou, & Mart´
ınez, 2007) two Slavic languages: Croatian (Kovacevic,
Jelaska, Kuvac, & Cepanec, 2005) and Slovene (Marjanoviˇ
c-Umek et al., 2008); one
Finno-Ugric language: Estonian (Tulviste, 2007); and one non-Indo-European language:
Basque (Barre˜
na, Ezeizabarrena, & Garc´
ıa, 2008; Barre˜
na et al., 2008).
The children acquiring a Germanic, or Slavic language as well as Estonian, or
French were predominantly monolinguals. About 20–25% of the Spanish participants, 30
(toddlers)–50% (infants) of the Basque participants, and 66% of the Galician participants
were bilingual. The bilingual children were assessed in their dominant language only.
Statistics
Effects of gender, age, and language community on the early use of gestures, as well
as on receptive and expressive vocabularies were explored by a multivariate analysis of
variance (MANOVA) and analyses of variance (ANOVAs) in full factorial models using SPSS
software. A logistic regression analysis was performed to explore the same independent
variables on whether a child had started to combine words. Age was scored in 3-month
intervals for both infants and toddlers. Because the number of CDI items varied between
versions, all analyses were recalculated with the percentage of the particular measure as
the dependent variable. The present authors’ comments are offered when this procedure
yields results differing from the analyses using the transformed raw data. Using raw
data assumes that each form is exhaustive, while using percentages assumes that each
form is equally exhaustive. Neither is correct and the truth lies somewhere in between.
Adaptations with longer CDI forms received higher scores by the first procedure, whereas
adaptations with shorter CDI forms, concentrating on the items expected to be most
prominent, are favoured by the second procedure. Means and standard deviations (SDs)
are given in raw scores.
As is often the case in field research, data did not meet all assumptions upon which
classic parametric tests such as ANOVA rest. The scales for word comprehension
and word production were transformed by the square root (word comprehension in
infants and word production in toddlers) and by a natural logarithm function, LN (word
production in infants) to approach normal distributions. The difference in number of
children from each language community has been handled by the use of the type III
332 M˚
arten Eriksson et al.
SS option in the ANOVAs. No deviance from homogeneity could be found for gender.
A level of significance of 1% was used to safeguard against type I errors. With these
precautions, and as the samples were unusually large, it was considered safe to calculate
MANOVAs and ANOVAs.
Because differences in distributions between groups are not revealed by statistics
based on means, analyses based on central tendencies will be complemented with
analyses of distributions for girls and boys separately. The distribution of scores for
girls versus scores for boys will be indexed using the variance ratio (VR), that is the
variance of the boys’ score over the variance of the girls’ score (Feingold, 1992). A VR of
1.00 indicates homogeneity of variance; a greater VR indicates greater variability among
boys and a VR less than 1.00 indicates greater variability among girls. The tail ratio (TR)
(Halpern, 2000) is another measure of variability that concentrates on the number of girls
and boys at the extreme ends of the distribution. It can be expressed as the percentage
of boys in the tail of all boys in the sample over the percentage of girls in the tail of
all girls in the sample. Thus, a TR over 1.00 indicates that boys are over-represented
in the particular tail and a TR under 1.00 indicates that girls are over-represented. An
equal distribution of girls and boys but with a difference in means results in an over-
representation of the low-performing group in the lower tail and an under-representation
of the same group in the upper tail. TRs, after adjustment for age, are also reported in
this study and the tails are set at the top and bottom 5%.
Results
Infants (W&G)
Communicative gestures, word comprehension, and word production were moderately
correlated with each other (rvaried between .38 and .59) and used as dependent
variables in a MANOVA with gender, age group, and language community as fixed
factors. The analysis revealed main effects of all factors. Girls (Wilks’ =.99, F[3,
4546] =9.895, p<.01, 2=.006) and older infants (Wilks’ =.55, F[6, 9092] =
534.886, p<.01, 2=.261) were reported to use and understand more types of
communicative items than boys and younger infants. The effect of language community
(Wilks’ =.76, F[24, 13185] =54.660, p<.01, 2=.09) showed that the infants’
language skills varied among communities. An interaction was found between age
and language community, indicating that the increase in language skills by age differed
among language communities (Wilks’ =.97, F[42, 13486] =3.267, p<.01, 2=.010).
No other interactions were found. Calculations using the percentage of mastered items,
comprehended as well as produced, relative to the number of items in the instrument
of each adaptation yielded the same results regarding significant contributions of the
independent variables, although all interactions now turned out significant.
First communicative gestures
A separate univariate ANOVA revealed main effects of age, gender, and language
community on children’s first communicative gestures. Girls (F[1, 4548] =26.79,
p<.01, 2=.006) were reported to use more types of communicative gestures than
boys and older infants were reported to use more types of communicative gestures
than younger infants (F[2, 4548] =1227.19, p<.01, 2=.35). See Table 2 for means
and SDs. The increase with age was slightly more pronounced among girls between
the first (8–10-month-olds) and second (11–13-month-olds) age groups. In contrast, the
Differences between girls and boys in early language 333
Ta b l e 2 . Mean (SD) number of first communicative gestures, word comprehension, and word
production for girls and boys (aged 0.08–1.4 [year; months]) and word production (aged 1.04–2.06) in
relation to age groups
Girls Boys
Age MSDNMSDN
First communicative gesture (W&G)
0.08–0.10 year 4.46 2.50 695 4.07 2.34 659
0.11–1.01 year 7.46 2.13 771 6.90 2.22 795
1.02–1.04 year 8.85 1.94 864 8.42 1.90 814
Word comprehension (W&G)
0.08–0.10 year 32.51 40.8 704 31.49 41.87 678
0.11–1.01 year 80.1 66.176 776 79.45 69.87 798
1.02–1.04 year 145.8 85.2 865 135.37 84.79 814
Word production (W&G)
0.08–0.10 year 1.5 3.14 704 1.46 3.43 678
0.11–1.01 year 8.0 13.19 776 6.39 11.00 798
1.02–1.04 year 24.5 36.83 901 18.89 29.14 834
Word production (W&S)
1.04–1.06 year 57.4 76.1 791 49.6 70.9 718
1.07–1.09 year 136.3 120.4 896 101.0 101.8 853
1.10–2.00 year 248.5 156.5 1132 203.7 147.5 1039
2.01–2.03 year 367.2 154.9 889 301.3 161.7 886
2.04–2.06 year 445.4 143.9 939 405.0 160.1 899
differences between girls and boys decreased at the oldest age group (14–16-month-olds),
most likely because of a ceiling effect for girls (F[2, 4548] =4.88, p<.01, 2=.002)
(Figure 1). There was a main effect of language community (F[8, 4548] =50.96, p<.01,
2=.08), and there was an interaction between age group and language community,
(F[14, 4585] =3.008, p<.01, 2=.009) indicating that children used more types of
gestures at an earlier age group in some language communities as compared with others.
There was no interaction between gender and language community. Calculations using
the proportion of communicative gestures yielded similar results (data not shown). The
VR ratio for communicative gestures was .96, indicating a very similar overall distribution
of scores for girls and boys. TR in the lower tail (5.4%) was 1.21, which means that there
0
2
4
6
8
10
0;08-0;10 0;11-1;01 1;02-1;04
Age (years; months)
Communicative gestures
girls
boys
Figure 1. Number of communicative gestures among girls and boys (aged 0.08–1.04) in relation to
age (N=4,598).
334 M˚
arten Eriksson et al.
0
5
10
15
20
25
30
0;08-0;10 0;11-1;01 1;02-1;04
Age (years; months)
Word produced
girls
boys
Figure 2. Number of words produced among girls and boys (aged 0.08–1.04) in relation to age (N=
4,598).
are 121 boys for every 100 girls in the lower tail of the distribution. TR among the top
8.1% was .69, indicating there are 69 boys for every 100 girls in this tail. Thus, despite an
extremely small effect size for gender, the difference between girls and boys on mean
number of communicative gestures had a clear impact on the tails.
Word comprehension
A corresponding ANOVA on word comprehension revealed main effects of age and
language community. Older infants (F[2, 4585] =1186.617, p<.01, 2=.34) were
reported to understand more words than younger infants and the number of words
children comprehended varied for different language communities (F[8, 4585] =82.278,
p<.01, 2=.12). See Table 2 for means and SDs. However, the effect of gender was
not statistically significant (p=.034); nor were there any interactions between gender
and age group or gender and language community, although there was an interaction
between age group and language community, (F[14, 4585] =2.825, p<.01, 2=.009)
indicating that children comprehended more words at an earlier age group in some
language communities as compared with others. Calculations that used the proportion of
words understood yielded similar results (data not shown). VR for word comprehension
was .93, with TR for the lower tail (5.1%) at 1.19 and TR for the upper tail (5%) at 1.03.
Thus, while VR based on means indicated a slightly greater variability among girls, the
analysis based on the tails gave the opposite result. As the differences are small, a fair
conclusion is that the variability between girls and boys in word comprehension is about
the same.
Word production
The same type of ANOVA revealed main effects of gender, age, and language community
on children’s early word production. Girls (F[1, 4641] =16.023, p<.01, 2=.003)
were reported to use more types of word than boys and older infants more types of
word than younger infants (F[2, 4641] =977.488, p<.01, 2=.296). See Figure 2 and
Table 2 for means and SDs. The effect of language community (F[8, 4641] =59.364,
p<.01, 2=.093) indicated that the number of words children from different language
communities produced varied. No interactions were significant. Calculations using the
proportion of word production yielded significant contributions of main effects as well
as for all interactions (data not shown). The VR for spoken words was .61, indicating
that girls were more variable than boys. However, because the sample was not normally
Differences between girls and boys in early language 335
0
50
100
150
200
250
300
350
400
450
500
1;04-1;06 1;07-1;09 1;10-2;00 2;01-2;03 2;04-2;06
Age (years, months)
Mean number of words produced
girls
boys
Figure 3. Mean number of words produced by girls and boys (aged 1.04–2.06) in relation to age (N=
9,012).
distributed, the VR value of .61 should be interpreted with caution. The proportion of
children from which no word was reported was 29%. The bottom TR was 1.13, indicating
slightly more boys in the bottom tail. About 5% of the children produced more than 43
words. The TR for these children was .65, indicating that there were considerably more
girls among the top 5%.
To sum up the effects of gender on infants’ communicative gestures, word compre-
hension, and word production as reported in the W&G form, the acquisition of all three
skills varied among language community and with the age of the children. Significant
differences were observed between girls and boys on communicative gestures and early
word production. There was one exception from a general girl advantage in word
production. Austrian boys produced more words than their female peers. We have no
explanation at hand for this but note that the sample of Austrian infant boys (42) was
the smallest group of all (Table 1). No significant difference was found between girls and
boys in comprehension. The greater male variability hypothesis received no support.
On the contrary, variability was comparable for both genders on the first two skills and
noticeably higher for girls in early word production. The TR analyses confirmed the
analyses based on mean estimates.
Toddlers (W&S)
Word production
An ANOVA that included gender, age group, and language community as fixed factors
revealed main effects of all three variables. Girls (F[1, 8942] =127.337, p<.01,
2=.014) were reported to use more types of word than boys and older toddlers
used more types of word than younger toddlers (F[4, 8942] =1633.699, p<.01, 2=
.422). See Table 2 for means and standard deviations. The difference between girls and
boys increased with age, (F[4, 8942] =3.895, p<.01, 2=.002, Figure 3). The effect
of language community (F[9, 8942] =36.600, p<.01, 2=.036) indicated that the
number of words children from different language communities produced varied. There
was also an interaction between age group and language community, (F[36, 8942] =
2.790, p<.01, 2=.011) indicating that children produced more words at an earlier age
336 M˚
arten Eriksson et al.
0
10
20
30
40
50
60
70
80
90
100
1;04-1;06 1;07-1;09 1;10-2;00 2;01-2;03 2;04-2;06
Age (years, months)
Mean percentage
girls
boys
Figure 4. Mean percentage of girls and boys reported to combine words (aged 1.04–2.06) in relation
to age (N=8,829 children, p<.05 for all age groups).
group in some language communities as compared with others. However, no interaction
between gender and language community was found. Calculations using the proportion
of words relative to the number of items in each adaptation of CDI did not change these
results. Girls from all language communities, including the Austrian toddlers, produced
now more words than the boys. The VR for these somewhat older children was .91,
indicating that the distribution was similar among girls and boys. As expected from the
difference in means, boys dominated the lower tail (4.9%, TR =1.28) and girls dominated
the upper tail (5%, TR =.70).
Onset of early two-word utterances
Differences in the onset of early two-word utterances between girls and boys were
investigated by parent reports of first-word combinations. These reports showed that
more girls (75.0%) combined words than boys (69.3%). These differences held across
all age groups but were most prominent between the ages 1.07 and 2.03, indicating
a ceiling effect in the oldest age group (Figure 4). A hierarchical logistic regression
analysis with age group, gender, and language community entered in a first block and
the interactions between age group and gender and between language community and
gender in a second block confirmed this difference. The analysis revealed main effects
for all variables in the first block (Wald statistics was 1812.98 for age group, 59.69 for
gender, and 136.08 for language community, B=1.19 and SE =.028 for age group and
B=.44 and SE =.058 for gender, all ps<.01). Girls had started to combine words to a
higher degree than boys. None of the interactions in the second block were significant.
In sum, there were significant differences between toddler girls and boys on word
production, differences that increased with age. Although word production varied
among children from different language communities, this did not affect the differences
between girls and boys. The distribution of spoken words was rather similar among girls
and boys and the tail ratios confirmed the analysis based on means. A similar but less
clear pattern was found for word combinations.
Gender differences across language communities
Previous analyses have shown main effects of language community for both infants (on
gestures, word comprehension, and word productions) and toddlers (word production
Differences between girls and boys in early language 337
Ta b l e 3 . Partial contribution (2) of gender in gender ×age group ANOVAs and the correlation
coefficient (rho) for each language community and the total
Infants Toddlers
Language Gestures Word Word Word Word
community (2) comprehension (2) production (2) production (2) combinations (rho)
Austria .010 .006 .004 .001 .052
Basque .025 .014 .002 .005 .064
Croatian .000 .001 .012 .034 .061
Danish .012 .000 .002 .025 .099
Estonian .013 .009 .041 .055 .138
French .030 .004 .013 .017 .110
Galician .003 .004 .001 .021 .064
Slovenian — .030 .118
Spanish .000 .000 .000 .020 .023
Swedish .029 .002 .002 .011 .044
Total .011 .001 .004 .020 .083
Note.‘’ indicates that boys scored higher than girls.
and onset of first word combinations). It is beyond the scope of the present study
to analyse the causes to these differences, some of which are likely to depend upon
variables we do not have control over. Instead, our interest concerned the issue whether
the emerging difference in early language skills between girls and boys differ between
language communities, disregarding the causes of these differences. Analyses so far have
not found any evidence of such interactions, although we have noticed a few deviant
cases where boys from one language community scored higher than the girls from
that particular community. Table 3 gives an overview of this pattern by showing the
effect sizes of the differences between girls and boys for each language community
and assessed language skill. Effect sizes (2) for gestures, word comprehension, and
word production were computed by a gender by age group ANOVA for each language
community (in contrast to the previous analyses in which language community was
included as a factor). Partial correlations were computed between gender and onset of
word combinations controlling for age. Analyses based on percentages did not result in
any essential change.
All effect sizes were small. Austrian and Galician infant boys deviated from the general
pattern and scored higher than infant Austrian and Galician girls on word production.
Galician boys scored also higher than Galician girls in word comprehension. Word
comprehension was the only language skill in which no general difference between girls
and boys was found. The girl advantage was without exceptions for toddlers and the
effect sizes were slightly larger for this group, although still small.
Discussion
The study has investigated gender differences in emerging language skills across 10
non-English language communities. Support was found for the hypotheses that girls are
generally ahead of boys in emerging language skills and that this difference increases with
age up to 2.06. Little support was found for the hypotheses that this gender difference
338 M˚
arten Eriksson et al.
in language varies across language communities or that young boys are more variable in
language skills than young girls.
Infant girls were reported to use more types of communicative gestures than infant
boys and both infant and toddler girls produced more words than same-aged boys.
Toddler girls were also ahead of boys in first word combinations. This finding is in accord
with that of other large-scale studies on English-speaking children (Bornstein et al.,
2004; Fenson et al., 2007; Galsworthy et al., 2000). In contrast to Fenson et al. (2007),
no difference between girls and boys was found in word comprehension. This absence
of a difference cannot be explained by floor or ceiling effects since comprehension
precedes production. Instead, we believe that the comprehension part of the CDI is
a less precise measure than the production part and, therefore, makes this difference
unstable and hard to assess. In fact, it was already suggested by Tomasello and Mervis
(1994) in the initial scientific publication on the CDI that parents are more uncertain
in judging their child’s comprehension as compared with production. If Tomasello and
Mervis are right, the uncertainty in parental reports on comprehension would leave more
room for parental expectations than reports on productions. However, as no difference
between girls and boys was found in comprehension, parental expectations do not seem
to be responsible for the differences between genders in gestures, word production,
and word combinations. In conclusion, Hypothesis 1 received support from data on
gestures; word production and onset of word combinations but not from the data on
word comprehension.
There were also main effects of language community in all five measures (for infants in
gestures, word comprehension, and word productions, for toddlers in word production
and word combinations), indicating a variation in the social and linguistic environments.
The variation in levels of language skills between language communities is important
because it indicates that the difference between girls and boys is robust. There was
no significant interaction between gender and language community. Thus, Hypothesis
2 received no support. On the contrary, no interaction between gender and language
community is precisely what we would expect if neuropsychological differences (e.g.,
differences in rate of brain lateralization or hormone release) were responsible for the
girl advantage in early language skills. However, a pattern of consistent differences
between girls and boys in language skills and consistent differences between language
communities but no interaction between gender and language community would be
compatible with social explanations to the girl advantage if the values and customs
related to gender roles are particularly robust with little variation between societies.
Examples would be if parents offered gender-typed toys (Fagot, 1995) or encouraged
more language towards their daughters (Leaper, Andersson, & Sanders, 1998) to exactly
the same extent in all the language communities investigated.
It was beyond the scope of the present study to collect data on gender norms in
different communities. However, many of the language communities in the present study
were also included in Hofstede’s (2001) comparison of values on several of dimensions
across nations. His masculinity index is of particular interest as it measures how distinct
gender roles are in one society as compared with another society. Slovakia was reported
with a top masculinity score of 110. Austria was also reported to score high on masculinity
(79), while gender roles in Denmark (16) and Sweden (5) were more or less overlapping.
This suggests that the studied language communities included substantial variation in
their distribution of gender roles.
It has been pointed out as a problem the fact that the population was not
homogeneous regarding demographic variables such as socioeconomic status (SES),
Differences between girls and boys in early language 339
birth order, or exposure to a second language. We agree that this would be a problem
if girls and boys were exposed differently to these factors. However, as girls and boys
were recruited in the same way within each community, and the sample was unusually
large, there is no reason to believe that these variables have favoured one sex over the
other. (The same assumption is regularly made for intelligence when not controlled for,
which it rarely is). Moreover, the only possible bias differences in SES and inclusion of
bilingualism could cause, would be to somewhat artificially increase the difference in
child performance between some language communities, not between genders. Large
differences between language communities would also increase the likelihood to obtain
significant interactions between language community and gender. Yet, we have not
focused on differences between language communities per se, but only intended to
investigate whether variation among language communities also affected the difference
between girls and boys. Furthermore, a main result in the study is that no interactions
between language community and gender were obtained. Hence, we cannot see that
incomplete control over SES, birth order, and inclusion of bilingual children has biased
the result in any interesting way. On the contrary, because bilingualism is so frequent
in the world, we think it is important to include such children in the study and show
that the girl advantage found for monolingual children also applies to bilingual children.
For number of communicative gestures and word production in both infants and
toddlers, the difference between girls and boys increased with age confirming Hypothesis
3. A ceiling effect, however, prevented the detection of this difference in first word
combinations. As neither girls nor boys speak at birth, any difference between genders
at a later point in time is expected to interact with age. However, interactions with age
have not been consistently reported in the literature. One reason for this inconsistency
is probably that age has been measured in small units such as months (e.g., Fenson et al.,
2007), which results in correspondingly more degrees of freedom between groups than
the present 3-month period and hence a lower F-value. How age is measured is naturally
a matter of convention but we stress that it should be categorized in a statistical model
in a way that gives expected effects a fair chance to be detected.
Because VR never was over 1, the greater male variability hypothesis (Hypothesis 4)
received no support. On the contrary, girls were consistently more variable than boys.
In particular, VR for spoken words among infants was .61. However, this difference in
variability can be explained by a floor effect among boys, something that is very clear for
word production among infants. A large proportion of the boys (30%) had not started to
use words productively at this age and VR for word production among toddlers increased
rapidly to .91. Except for word comprehension, the TR analyses showed that boys were
consistently over-represented at the lower end of the distributions and that girls were
over-represented at the top. Taken together, this indicates basically uniform distributions
with a small difference in means between girls and boys in gestures and spoken words
and confirms the analyses based on means. Equal variance in language skills for girls
and boys below the age of 3 years is compatible with the findings of both Arden and
Plomin (2006) and the review of Feingold (1992). However, the VR for older children
would have to support one or the other indicating either that a greater male variability
in language skills emerges from the age of 3 in accord with general intelligence (Arden
& Plomin, 2006), or that the variance between genders in language skills departs from
that found for general intelligence and remains equal (Feingold, 1992).
Merging of actual data in a re-analysis has the advantage over conventional meta-
analyses based on effect size in that it allows not only the investigation of main effects but
also interactions between variables and exploration of distributions. However, analysing
340 M˚
arten Eriksson et al.
already collected data also imply certain drawbacks such as a lack of control over
sample sizes, of the applied instruments and in choosing what background information
to include. Another limitation of the present study, shared with all parent report
instruments, is that it is based on indirect measures of children’s performance. However,
parent reports on early language skills have proved to be valid in several studies (see
Fenson et al., 2007 for a review) and are particularly useful for depicting emergent
phenomena. Unique features of the present study are its comprehensiveness in including
over 13,000 girls and boys from 10 different language communities assessed with
adaptations of the same instrument.
It must be noted that the effect of gender on parent assessment of infants’ and
toddlers’ language skills was small (gender accounted for around 1% or less of the
variability in children’s early language skills). Such small effect sizes are often said to be
of no importance (e.g., Hyde, 2005). However, the difference between girls and boys
corresponded to .7 gesture and 1 spoken word at 1.00 and 57 spoken words at 2.00. At
least at the age of 2.00, a difference of this size have important implications for issues
of screening for language delay. It indicates either that one gender is more vulnerable
and therefore more in need for special interventions and unisex norms should be kept,
or that unisex norms would result in more false negative girls and more false positive
boys as compared with a situation in which different norms are used. That would make
the screen costly and be disadvantageous towards girls in the sense that it is worse to
be in need of help but not receive it than to receive help without needing it. Separate
long-term follow-ups of girls and boys at the lower tail would be needed to delineate
these alternatives.
Acknowledgements
We would like to express our sincere gratitude to all participants of all countries and to our
colleagues of the European Network on Communicative Development Inventories for their
support and input during our meetings. Parts of this study were supported by the Austrian
Science Fund (FWF; P19581-B02), the City of Graz, the Lanyar Foundation (P325), the Croatian
National Grant (MZOS 0013 002), and the Estonian Research Competency Council (grant No.
SF0180025s08). The study has also received financial support from the University of G¨
avle.
References
Arden, R., & Plomin, R. (2006). Sex differences in variance of intelligence across childhood.
Personality and Individual Differences,41, 39–48. doi:10.1016/j.paid.2005.11.027
Barre˜
na, A., Ezeizabarrena, Ma. J., & Garc´
ıa, I. (2008). Influence of the linguistic environment on
the development of the lexicon and grammar of Basque bilingual children. In C. P´
erez-Vidal,
M. Juan-Garau, & A. Bel (Eds.), A portrait of the young in the new multilingual Spain. Issues
in the acquisition of two or more languages in multilingual environments (pp. 86–110).
Multilingual Matters.
Barre˜
na, A., Garc´
ıa, I., Ezeizabarrena, Ma. J., Almgren, M., Arratibel, N., Olano, I., . . . Colina, A.
(2008). MacArthur-Bates komunikazio garapena neurtzeko zerrenda. Euskarara egokitua.
Erabiltzaileentzako gida eta eskuliburu teknikoa. [The MacArthur-Bates Communicative
Development Inventory. Adapted to Basque. User’s manual and technical handbook.] Bilbo:
Udako Euskal Unibertsitatea/Bilbao, Spain: Basque Summer University.
Bates, E., Benigni, L., Bretherton, I., Camaioni, L., & Volterra, V. (1979). The emergence of symbols:
Cognition and communication in infancy. New York: Academic Press.
Differences between girls and boys in early language 341
Berglund, E., & Eriksson, M. (2000). Communicative development in Swedish children 16–28
months old. The Swedish early communicative development inventory – words and sentences.
Scandinavian Journal of Psychology,41(2), 133–144. doi:10.1111/1467-9450.00181
Best, D. L., & Williams, J. E. (1997). Sex, gender and culture. In J. W. Berry, M. H. Segall, &
C. Kagitcibasi (Eds.), Handbook of cross-cultural psychology: Vol 3. Social behavior and
applications (pp. 163–212). Boston: Allyn and Bacon.
Bleses, D., Vach, W., Slott, M., Wehberg, S., Thomsen, P., Madsen, T., . . . Basbøll, H. (2008). The
Danish Communicative Development Inventories: Validity and main developmental trends.
Journal of Child Language,35, 619–650.
Bleses, D., Vach, W., Wehberg, S., Faber, K., & Madsen, T. O. (2007). Tidlig kom-
munikativ udvikling. Et værktøj til beskrivelse af sprogtilegnelse baseret p˚
aCDI-
forældrerapportundersøgelser af danske normalthørende og hørehæmmede børn. [Early
communicative development. A tool for the description of language acquisition based on CDI
parental reports for normal hearing and hearing impaired Danish children.] Odense: Syddansk
Universitetsforlag.
Bornstein, M. (2002). Toward a multiculture, multiage, multimethod science. Human Develop-
ment,45, 257–263. doi:10.1159/000064986
Bornstein, M., & Cote, L. (2005). Expressive vocabulary in language learners from two ecological
settings in three language communities. Infancy,7, 299–316. doi:10.1207/s15327078in0703 5
Bornstein, M., Cote, L., Maital, S., Painter, K., Park, S. -Y., Pascual, L., . . . Vyt, A. (2004). Cross-
Linguistic analysis of vocabulary in young children: Spanish, Dutch, French, Hebrew, Italian,
Korean, and American English. Child Development,75, 1115–1139. doi:10.1111/j.1467-
8624.2004.00729.x
Bornstein, M., Hahn, C. -S., & Haynes, O. (2004). Specific and general language performance
across early childhood: Stability and gender considerations. First Language,24(3), 267–304.
doi:10.1177/0142723704045681
Bornstein, M., Haynes, O. M., Pascual, L., Painter, K., & Galperin, C. (1999). Play in two
societies: Pervasiveness of process, specificity of structure. Child Development,70, 317–331.
doi:10.1111/1467-8624.00024
Caldera, Y., Huston, A., & O‘Brian, M. (1989). Social interactions and play patterns of parents
and toddlers with feminine, masculine and neutral toys. Child Development,60, 70–76.
doi:10.2307/1131072
Crow, T. (1998). Sexual selection, timing and the descent of man: A theory of the genetic origins
of language. Cahiers de Psychologie Cognitive/Current Psychology of Cognition,17, 1079–
1114.
Eagly, A., Wood, W., & Diekman, A. (2000). Social role theory of sex differences and similarities:
A current appraisal. In T. Eckes, & H. Trautner (Eds.), The development of social psychology
of gender (pp. 123–174). Mahwah, NJ: Erlbaum.
Ellis, H. (1894). Man and woman. London: Walter Scott.
Eriksson, M., & Berglund, E. (1999). Swedish early communicative development inventories: Words
and gestures. First Language,19(55), 55–90.
Fagot, B. (1995). Parenting boys and girls. In M. Bornstein (Ed.), Handbook of parenting. Vol. 1:
Children and parenting (Vol. 1, pp. 163–183). Mahwah, NJ: Lawrence Erlbaum.
Feingold, A. (1992). Sex differences in variability in intellectual abilities: A new look at an old
controversy. Review of Educational Research,62, 61–84.
Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., & Pethick, S. J. (1994). Variability in early
communicative development. Monographs of the Society for Research in Child Development,
59, 1–189. doi:10.2307/1166093
Fenson, L., Marchman, V. A., Thal, D. J., Dale, P., Reznick, S., & Bates, E. (2007). MacArthur-Bates
communicative development inventories (2nd ed.). Baltimore: Brookes Publishing.
Gallego, C. & Mariscal, S. (2008). Psychosocial traits of children identified as at risk for
language delay by the Spanish McArthur-CDI. The Second European Network Meeting on
Communicative Development Inventories. University of Zagreb, Dubrovnik, Croatia.
342 M˚
arten Eriksson et al.
Galsworthy, M., Dionne, G., Dale, P., & Plomin, R. (2000). Sex differences in early verbal and
non-verbal cognitive development. Developmental Science,3, 206–215. doi:10.1111/1467-
7687.00114
Garc´
ıa, I., Ezeizabarrena, Ma. J., Almgren, M., & Errarte, I. (2005). “El desarrollo l´
exico infantil
en euskara: Una nueva adaptaci´
on del test MacArthur”. [The development of child lexicon
in Basque: a new adaptation of the Mac Arthur test.]. In Mayor Cinca, Zubiauz de Pedro, &
Diez-Villoria (Eds.), Estudios sobre la adquisici´
on del lenguaje (pp. 955–966). [Studies on
language acquisition] Ediciones Universidad de Salamanca, Salamanca
Halpern, D. (2000). Sex differences in cognitive abilities (3rd ed.). Mahwah, NJ: Lawrence Erlbaum
Associates.
Hofstede, G. (2001). Culture’s consequences: Comparing values, behaviors, institutions, and
organizations across nations. Thousand Oaks, CA: Sage Publications.
Hyde, J. (2005). The gender similarities hypothesis. American Psychologist,60(6), 581–592.
doi:10.1037/0003-066X.60.6.581
Hyde, J., & Linn, M. (1988). Gender differences in verbal ability: A meta-analysis. Psychological
Bulletin,104, 53–69. doi:10.1037/0033-2909.104.1.53
Jacklin, C. N., & Baker, L. (1993). Early gender development. In S. Oskamp, & M. Costanzo (Eds.),
Gender issues in contemporary society (Vol. 6, pp. 41–57). Newbury Park, CA: Sage.
Jackson-Maldonado, D., Thal, D., Bates, E., Marchman, V., & Gutierrez-Clellen, V. (1993). Early
lexical development of Spanish-speaking infants and toddlers. Journal of Child Language,
20(3), 523–549. doi:10.1017/S0305000900008461
Kern, S. (2007). Lexicon development in French-speaking infants. First Language,27, 227–250.
doi:10.1177/0142723706075789
Kern, S., & Gayraud, F. (2007). Influence of preterm birth on early lexical and grammatical
acquisitions in French. First Language,27, 159–173. doi:10.1177/0142723706075789
Kovacevic, M., Jelaska, Z., Kuvac, J., Cepanec, M. (2005). Komunikacijska razvojna ljestvica
KORALJE (Communicative Development Inventories). Second version. Jastrebarsko: Naklada
Slap.
Law, J., Boyle, J., Harris, F., Harkness, A., & Nye, C. (1998). Screening for speech and language
delay: A systematic review of the literature. Health Technology Assessment,2(9).
Leaper, C., Andersson, K., & Sanders, P. (1998). Moderators of gender effects on parents’ talk
to their children: A meta-analysis. Developmental Psychology,34, 3–27. doi:10.1037/0012-
1649.34.1.3
L´
opez Ornat, S., Gallego, C., Gallo, P., Karousou, A., Mariscal, S. & Mart´
ınez, M. (2005).
Inventario de Desarrollo Comunicativo MacArthur. Adaptaci´
on espa˜
nola. [The MacArthur
Communicative Development Inventory. Adaptation to Spanish.]: Madrid: TEA.
Lyytinen, P., Poikkeus, A. M., & Laakso, M. L. (1997). Language and symbolic play
in toddlers. International Journal of Behavioral Development,21(2), 289–302.
doi:10.1080/016502597384875
Maccoby, E. E., & Jacklin, C. N. (1974). The psychology of sex differences. Stanford, CA: Stanford
University Press.
Mariscal, S., L´
opez Ornat, S., Gallego, C., Karousou, A., & Mart´
ınez, M. (2007). La evaluaci´
on
del desarrollo comunicativo y ling¨
u´
ıstico mediante la versi´
on espa˜
nola de los inventarios
MacArthur-Bates. [The assessment of communicative and linguistic development applying the
Spanish version of the MacArthur-Bates Inventories]. Psicothema,19(2), 190–197.
Marjanoviˇ
c-Umek, L., Fekonja, U., Kranjc, S., & Bajc, K. (2008). The effect of children’s gender and
parental education on toddler language development. European Early Childhood Education
Research Journal,16, 325–342. doi:10.1080/03004430600851280
Marschik, P. B., Einspieler, C., Garzarolli, B., & Prechtl, H. F. R. (2007). Events at
early development: Are they associated with early word production and neurodevel-
opmental abilities at the preschool age? Early Human Development,83, 107–114.
doi:10.1016/j.earlhumdev.2006.05.009
Differences between girls and boys in early language 343
P´
erez-Pereira, M. (2008). Early Galician/Spanish bilingualism: Contrasts with monolingualism. In
C. P´
erez-Vidal, M. Juan-Garau, & A. Bel (Eds.), A portrait of the young in the new multilingual
Spain (pp. 39–62). Clevedon: Multilingual Matters.
P´
erez-Pereira, M., & Garc´
ıa-Soto, X. R. (2003). El diagn´
ostico del desarrollo comunicativo en
la primera infancia: Adaptaci´
on de las escalas MacArthur al gallego. [The diagnosis of
communicative development in early childhood. Adaptation of the MacArthur scales to
Galician.]. Psicothema,15(3), 352–361.
P´
erez Pereira, M., & Resches, M. (2011). Concurrent and predictive validity of the Galician CDI.
Journal of Child Language, 38(1), 121–140. doi:10.1017/S0305000909990262
Suizzo, M. -A., & Bornstein, M. (2006). French and European American child-mother play: Culture
and gender considerations. International Journal of Behavioral Development,30(6), 498–
508. doi:10.1177/0165025406071912
Tomasello, M., & Mervis, C. B. (1994). The instrument is great, but measuring comprehension is
still a problem. Monographs of the Society for Research in Child Development,59, 174–179.
Tulviste, T. (2007). Variation in vocabulary development among Estonian children as a function of
child’s gender, birth order, child-care, and parental education. In M. Eriksson (Ed.), Proceedings
from the First European Network Meeting on the Communicative Development Inventories
(pp. 16–21). G¨
avle, Sweden: University of G¨
avle.
Received 4 October 2010; revised version received 26 April 2011
... Numerous studies have reported that gender, parental education, and kindergarten attendance have an effect on children's language development (Huttenlocher et al., 1991;Arriaga et al., 1998;Bornstein and Haynes, 1998;Feldman et al., 2000;D'Odorico et al., 2001;Bauer et al., 2002;Bavin et al., 2008;Umek et al., 2008;Eriksson et al., 2012;Hammer et al., 2017). Bouchard et al. (2009) used the Quebec French version of the MacArthur Communicative Developmental Inventories (Trudeau et al., 1999) to investigate gender differences in language development in French Canadian children between 8 and 30 months of age. ...
Article
Full-text available
The aim of this study was to evaluate the psychometric properties of the adapted Cyprus Greek Lexical List a-CYLEX (GR) in a sample of 194 Greek toddlers from the island of Crete with Standard Modern Greek (SMG) as their primary language. The a-CYLEX (GR) is a parental report checklist for assessing the receptive and expressive vocabulary skills of children aged 12 months to 3:6 years. Concurrent validity of the instrument was tested via correlations with the adapted Greek version of the Receptive One-Word Picture Vocabulary Test-II (ROWPVT-II), which was administered to 124 SMG-speaking children between the ages of 2 and 3:6 years. Test–retest reliability was tested by administering the instrument two times within a 2-week interval to 59 parents (30.41% of the total sample). Statistical analyses provided strong evidence for the high internal consistency and test–retest reliability of the a-CYLEX (GR). The role of the demographic variables in vocabulary performance and the frequency of each a-CYLEX (GR) word category by age were also investigated. In conclusion, the a-CYLEX (GR) is a parental report checklist that can be used by clinicians who are interested in assessing receptive and expressive vocabulary of children during toddlerhood.
... Finally, variability may be associated with several measured and unmeasured additional factors. For example, at 12 months in the current study, higher CBR was associated with: female sex (language is often reported to be more advanced in females than males (e.g., [3,16,20,23]), though this is confounded in by the sex imbalance in the HR-ASD group); higher maternal education (which has been associated with more advanced language development (e.g., [15,34])) and site, which may relate to minor administration differences or underlying demographic differences between sites. ...
Article
Full-text available
Background Canonical babbling—producing syllables with a mature consonant, full vowel, and smooth transition—is an important developmental milestone that typically occurs in the first year of life. Some studies indicate delayed or reduced canonical babbling in infants at high familial likelihood for autism spectrum disorder (ASD) or who later receive an ASD diagnosis, but evidence is mixed. More refined characterization of babbling in the first year of life in infants with high likelihood for ASD is needed. Methods Vocalizations produced at 6 and 12 months by infants (n = 267) taking part in a longitudinal study were coded for canonical and non-canonical syllables. Infants were categorized as low familial likelihood (LL), high familial likelihood diagnosed with ASD at 24 months (HL-ASD) or not diagnosed (HL-Neg). Language delay was assessed based on 24-month expressive and receptive language scores. Canonical babble ratio (CBR) was calculated by dividing the number of canonical syllables by the number of total syllables. Generalized linear (mixed) models were used to assess the relationship between group membership and CBR, controlling for site, sex, and maternal education. Logistic regression was used to assess whether canonical babbling ratios at 6 and 12 months predict 24-month diagnostic outcome. Results No diagnostic group differences in CBR were detected at 6 months, but HL-ASD infants produced significantly lower CBR than both the HL-Neg and LL groups at 12 months. HL-Neg infants with language delay also showed reduced CBR at 12 months. Neither 6- nor 12-month CBR was significant predictors of 24-month diagnostic outcome (ASD versus no ASD) in logistic regression. Limitations Small numbers of vocalizations produced by infants at 6 months may limit the reliability of CBR estimates. It is not known if results generalize to infants who are not at high familial likelihood, or infants from more diverse racial and socioeconomic backgrounds. Conclusions Lower canonical babbling ratios are apparent by the end of the first year of life in ASD regardless of later language delay, but are also observed for infants with later language delay without ASD. Canonical babbling may lack specificity as an early marker when used on its own.
... The 27-to 30-month-age bin shows a slightly weaker association, but with small sample sizes (< 30) for some age bins, this may be due to sampling variability. Figure D1 in Appendix D shows that the relationships between children's age and (a) their total vocabulary from the full CDI:WS (left), (b) their full CDI ability (middle), and (c) their ability measured from the CDI-CAT (right) are all of approximately the same strength and show the modest female advantage commonly found in assessments of young children's language ability (Eriksson et al., 2012; see the work of Frank et al., 2021, Ch. 6, for an overview). ...
Article
Full-text available
Purpose Measuring the growth of young children's vocabulary is important for researchers seeking to understand language learning as well as for clinicians aiming to identify early deficits. The MacArthur–Bates Communicative Development Inventories (CDIs) are parent report instruments that offer a reliable and valid method for measuring early productive and receptive vocabulary across a number of languages. CDI forms typically include hundreds of words, however, and so the burden of completion is significant. We address this limitation by building on previous work using item response theory (IRT) models to create computer adaptive test (CAT) versions of the CDIs. We created CDI-CATs for both comprehension and production vocabulary, for both American English and Mexican Spanish. Method Using a data set of 7,633 English-speaking children ages 12–36 months and 1,692 Spanish-speaking children ages 12–30 months, across three CDI forms (Words & Gestures, Words & Sentences, and CDI-III), we found that a 2-parameter logistic IRT model fits well for a majority of the 680 pooled vocabulary items. We conducted CAT simulations on this data set, assessing simulated tests of varying length (25–400 items). Results Even very short CATs recovered participant abilities very well with little bias across ages. An empirical validation study with N = 204 children ages 15–36 months showed a correlation of r = .92 between language ability estimated from full CDI versus CDI-CAT forms. Conclusion We provide our item bank along with fitted parameters and other details, offer recommendations for how to construct CDI-CATs in new languages, and suggest when this type of assessment may or may not be appropriate.
... Similarly, there was a pragmatic growth with age found in Italian children's data [31] and in the speech of Norwegian children [31]. Whereas past researchers have found gender differences in language development where girls perform better than boys [32][33][34][35][36][37], the present study has shown that gender seems to have minor or zero effect on pragmatic language skills of preschoolers in the test's Arabian context. This finding is consistent with several studies in the literature where there is no gender effect [31,[38][39][40]. ...
Article
Full-text available
The scope of this study is threefold. First, it examines socioeconomic status (SES) andpragmatic language development (PLD), hypothesizing that parental education and employmentlevels are mediators, where SES affects PLD mainly through its influence on parental educationand employment levels. Second, we used quantitative (age) and qualitative (gender) variablesas moderators, hypothesizing that gender and age moderate the relationship between classroominteraction (CI), social interaction (SI), and personal interaction (PI) skills and level of PLD. Third,characteristics of PLD in preschoolers with and without pragmatic language impairment (PLI) arecompared. The Arabic Pragmatic Language Skills Inventory (A-PLSI) was administered by preschoolteachers and speech-language pathologists in Saudi Arabia to 264 preschoolers with and withoutneurodevelopmental disorders to assess their PLD. Additionally, the SES of the parents of the samenumber of participants was surveyed. Results show that the more CI, SI, and PI a preschool childhas, the greater the likelihood of typical PLD, regardless of the parents’ employment or education.Further, we obtained evidence that CI, SI, and PI all grow together with age. However, these threeelements (namely, CI, SI, and PI) do not seem to be moderated or altered by gender. Typical PLD may be attainable when preschool children demonstrate typical mental and physical development, incontrast to children with psychiatric histories who display atypical PLD. These findings suggest that preschool children with more CI, SI, and PI will exhibit more typical PLD. The achievement of thisgoal results in a sustainable society for children
Article
Full-text available
The parents' attention is considerably important during the developmental period of children. It is a matter of concern that parents remain busy in their employment and daily routine, where they have no time to interact with the children. Hence to compensate parents' unavailability children use screen to pass their time which is drastically affecting their language skills. The present study investigates the "Relationship between parental negligence, screen time and expressive language delay among young children". Three Questionnaires i.e. Parental Acceptance-Rejection/Control Questionnaire short Form, Language Use Inventory and Child Technology Test are used to collect data from 150 mothers of young children 1 ½ to 4 ½ years who were visiting speech pathology rehabilitation centres of Islamabad. A half portion of the sample is collected via online due to pandemic COVID-19, and half by direct interaction when lockdown partly subsided. The convenient sampling is used. The study shows significant relationship between parental negligence and screen time but no relationship found between parental negligence and expressive language delay among young children. Further, it concludes that screen duration remained high in the young male children than in females while the expressive language delay is more prevalent in the children belonging to nuclear families.
Article
Full-text available
Over the past few decades, several studies have focused on potential sex-related differences in the trajectories of language development and functioning. From a behavioral point of view, the available literature shows controversial results: differences between males and females in language production tasks may not always be detectable and, even when they are, are potentially biased by sociological and educational confounding factors. The problem regarding potential sex-related differences in language production has also been investigated at the neural level, again with controversial results. The current minireview focuses on studies assessing sex-related differences in the neural networks of language production. After providing a theoretical framework of language production, it is shown that the few available investigations have provided mixed results. The major reasons for discrepant findings are discussed with theoretical and methodological implications for future studies.
Article
This study explored associations between mothers’ language teaching practices and children’s language skills concurrently and longitudinally, while also taking into account the children’s sex and mothers’ education. Estonian mothers of 76 children reported their language teaching practices at child ages 3;0 and 4;0. Children’s language comprehension and production were measured via the examiner-administered New Reynell Developmental Language Scales (NRDLS). The results indicated that at Wave 1, girls scored higher on language comprehension and production than boys. Mothers’ higher education predicted higher scores on language production. Maternal corrective feedback was a negative predictor of children’s concurrent language comprehension. At Wave 2, neither maternal teaching practices, maternal education nor child’s sex predicted the language measures of interest. The longitudinal results showed that later language production scores were negatively predicted by mothers’ corrective feedback. Yet, the strongest predictors of both language scores were the scores of language comprehension and production measured 1 year earlier. Moreover, maternal language teaching practices at Wave 2 were predicted by mothers’ early language teaching practices, but not by children’s earlier language skills.
Article
The mathematics achievement discrepancy between girls and boys, with its subsequent occupational consequences, is an issue that has received considerable attention in the literature. It is often referred to as the ‘math‐gap’ and favours boys. A major component of the explanation of this gap resides in determining its age of onset. We analyse here data from more than 10,000 (cross‐sectional study) and 2000 (longitudinal study) French students aged 4–7 years, tested in the framework of the Etude Longitudinale Française depuis l'Enfance (ELFE). The results allow to precisely determine the age of onset, since the gender difference, non‐existent (or even slightly in favour of girls) in kindergarten (4–5 years), is clearly in favour of boys in first grade (6–7 years). They could therefore provide an important element in the controversial debate on the origin of gender‐differentiated performance in mathematics.
Article
Gesture is an important communication tool that provides insight into infants' early language and cognitive development and predicts later language skills. While bilingual school‐age children have been reported to gesture more than monolinguals, there is a lack of research examining gesture use in infants exposed to more than one language. In this preregistered study, we compared three groups of 14‐month‐old infants (N = 150) learning French and/or English: bilinguals (hearing a second language at least 25% of the time), exposed (hearing a second language 10%–24% of the time), and monolinguals (hearing one language 90% of the time or more). Parent‐reported use of communicative gestures was gathered from the MacArthur‐Bates Communicative Development Inventories (CDI). Results showed that the three language groups had similarly sized gesture repertoires, suggesting that language exposure did not affect gesture development at this age. However, a gender effect was found, where girls produced more types of gestures than boys. Overall, these results suggest that gender, but not language exposure, contributes to differences in gesture development in infancy.
Article
Full-text available
The MacArthur Communicative Development Inventories are a valid and reliable instrument for assessing young children's communicative and linguistic development. The present paper offers data coming from the adaptation of the MacArthur scales to Galician and their application to a sample of 722 Galician-speaking children between 8 and 30 months of age. Means and standard deviations scores for the different parts of the scale are offered. An analysis of the results obtained is carried out. An initial analysis is offered of the type of vocabulary used at different points in development. Finally data on reliability of the scales are given.
Article
This study compared multiple characteristics of girls' and boys' vocabulary in 6 different linguistic communities—1 urban and 1 rural setting in each of 3 countries. Two hundred fifty-two mothers in Argentina, Italy, and the United States completed vocabulary checklists for their 20-month-old children. Individual variability was substantial within each linguistic community. Minimal cross-linguistic differences were found in children's vocabulary size; however, differences among languages in the composition of children's vocabularies appeared possibly related to cultural valuing of different categories of words. Ecological setting differences within cultures appeared in children's vocabulary size, even when the composition of children's vocabularies was examined: Children living in urban areas were reported by their mothers to say significantly more words than children living in rural areas, particularly for Argentine and U.S. children. Girls had consistently larger vocabularies than boys. These findings are discussed in terms of contextual and child factors that together influence first language learning.
Article
Contemporary research on sex differences in intellectual abilities has focused on male-female differences in average performance, implicitly assuming homogeneity of variance. To examine the validity of that assumption, this article examined sex differences in variability on the national norms of several standardized test batteries. Males were consistently more variable than females in quantitative reasoning, spatial visualization, spelling, and general knowledge. Because these sex differences in variability were coupled with corresponding sex differences in means, it was demonstrated that sex differences in variability and sex differences in central tendency have to be considered together to form correct conclusions about the magnitude of cognitive gender differences.
Article
Altogether 329 children participated in four longitudinal studies of specific and general language performance cumulatively from 1;1 to 6;10. Data were drawn from age-appropriate maternal questionnaires, maternal interviews, teacher reports, experimenter assessments and transcripts of children’s own spontaneous speech. Language performance at each age and stability of individual differences across age in girls and boys were assessed separately and together. Across age, including the important transition from preschool to school, across multiple tests at each age and across multiple reporters, children showed moderate to strong stability of individual differences; girls and boys alike were stable. In the second through fifth years, but not before or after, girls consistently outperformed boys in multiple specific and general measures of language.