Early Childhood Research Quarterly 36 (2016) 259–268

Contents lists available at ScienceDirect

Early Childhood Research Quarterly

Mathematics and language: Individual and group differences in

mathematical language skills in young children
David J. Purpura a,∗ , Erin E. Reid b
a
Purdue University, IN, United States
b
Erikson Institute, IL, United States

a r t i c l e i n f o a b s t r a c t

Article history: The development of early numeracy knowledge is influenced by a number of non-mathematical
Received 16 June 2015 factors—particularly language skills. However, much of the focus on the relation between language and
Received in revised form early numeracy has utilized general language measures and not domain-specific measures of mathemat-
21 December 2015
ical language. The primary purpose of this study was to determine if the variance accounted for by general
Accepted 27 December 2015
language skills in predicting numeracy performance was better accounted for by mathematical language.
Available online 15 January 2016
Further, age- and parental education-related differences in mathematical language performance were
explored. Using a sample of 136 3- to 5-year-old preschool and kindergarten children (M = 4.28 years,
Keywords:
Mathematical language
SD = 0.67 years), a series of mixed-effect regressions were conducted. Results indicated that although
Numeracy general language performance was initially a significant predictor of numeracy performance, when both
Preschool mathematical language and general language were included in the model, only mathematical language
Kindergarten was a significant predictor of numeracy performance. Further, group-difference analyses revealed that
Math children from families where both parents had less than a college education performed significantly
lower on mathematical language than their peers; and even by 3-years-old, children have acquired a
substantial body of mathematical language skills. Implications and future directions are discussed.
© 2015 Elsevier Inc. All rights reserved.

1. Introduction
Developing early numerical competencies is a complex and con-
tinual process that begins even prior to formal schooling (Sarama
& Clements, 2009; Starkey, Klein, & Wakeley, 2004). Unfortu-
nately, many children, particularly those from families of low
socio-economic-status (SES) struggle to acquire basic numeracy
skills and remain behind their peers from families of middle- and
high-SES (Jordan, Huttenlocher, & Levine, 1994; Jordan & Levine,
2009; Starkey et al., 2004). It is clear that the developmental
process of numeracy skill acquisition does not occur indepen-
dently from other academic and cognitive skills. For example,
LeFevre et al. (2010) identified three specific pathways by which
children typically acquire early numeracy knowledge: linguistic
(Hooper, Roberts, Sideris, Burchinal, & Zeisel, 2010; Juel, 1988),
spatial attention (or executive functioning; McClelland, Acock, &
Morrison, 2006), and quantitative (or the approximate number
system; Dehaene, 1997; Halberda, Mazzocco, & Feigenson, 2008).
These three pathways were based on the three modules identified
in Dehaene’s triple-code model of number processing (Dehaene,
1992). Although LeFevre et al. noted that each of these pathways
uniquely contributes to numeracy knowledge, the linguistic path-
way was the most consistent and strongest predictor of early
numeracy.
Importantly, the focus of the linguistic pathway is often placed
on general language skills/processing (e.g., general vocabulary,
phonological awareness), but there is evidence that the language
associated with numeracy is highly content-specific (mathemat-
ical language; Harmon, Hedrick, & Wood, 2005). However, there
is a limited depth of research examining the nature of the lan-
guage associated with numeracy skills, particularly prior to entry
to formal schooling. The central goal in this study was to evaluate
individual and group differences in an understudied aspect of the
linguistic pathway – children’s knowledge of the words utilized in
early mathematics, or mathematical language – and the relation of
these differences to general numeracy performance.

1.1. The importance of language for numeracy development

∗ Corresponding author at: Purdue University, Human Development and Family
Studies, 1202W State St. Rm. 231, West Lafayette, IN 47907, United States. The strong relation between numeracy and language is evident
E-mail address: purpura@purdue.edu (D.J. Purpura). relatively early in children’s academic development. In preschool

http://dx.doi.org/10.1016/j.ecresq.2015.12.020
0885-2006/© 2015 Elsevier Inc. All rights reserved.
260 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268
and early elementary school, children’s language skills and numer-
acy ability are highly related (Hooper et al., 2010; Purpura, Hume,
Sims, & Lonigan, 2011; Romano, Babchishin, Pagani, & Kohen,
2010). This relation appears to be general to nearly all aspects of
early symbolic numeracy and not specific to individual compo-
nents of the symbolic system, as Purpura and Ganley (2014) found
that language skills accounted for significant variance in predicting
nearly all early numeracy skills. However, these connections are
presumed to begin even earlier than preschool as substantial evi-
dence indicates that numeracy and language processing activate
similar areas of the brain—suggesting they share some common
neural pathways (Baldo & Dronkers, 2007; Dehaene, Spelke, Pinel,
Stanscu, & Tsivkin, 1999).
Language is the central feature that serves to distinguish the
symbolic numeracy system (which typically begins to develop
even as young as 18–24 months; Mix, 2009; Sarama & Clements,
2009) from the more primitive non-symbolic approximate quantity
discrimination system which is believed to be present beginning
in infancy (Dehaene et al., 1999; Spelke & Tsivkin, 2001; Starr,
Libertus, & Brannon, 2013; Xu & Spelke, 2000). As children get
older, difficulties in numeracy often co-occur with difficulties in
literacy and language (Lewis, Hitch, & Walker, 1994; Mann Koepke
& Miller, 2013). Those children with difficulties in both areas tend
to have more severe (Hanich, Jordan, Kaplan, & Dick, 2001; Jordan
& Hanich, 2000) and persistent (Silver, Pennet, Black, Fair, & Balise,
1999) numeracy difficulties than children with numeracy difficul-
ties alone.
Conceptually, numeracy and language are linked in develop-
ment because language skills are believed to support numerical
development (Miura & Okamoto, 2003) as children utilize language
skills – particularly number words – to refine their quantitative
understanding (Spelke, 2003). Findings from cross-cultural studies
suggest that individuals who speak languages that do not have a
well-developed counting system are limited in their ability to grasp
exact quantities (Gordon, 2004). For example, in languages that
have plural markers (e.g., English and Russian), children more read-
ily acquire the concept of “two = ••” than in languages that do not
have plural markers (e.g., Chinese and Japanese; Li, Sun, Baroody, &
Purpura, 2013; Sarnecka, Kamenskaya, Yamana, Ogura, & Yudovina,
2007). Even within English, children must refine their concepts of
quantity through the use of number words. Often, the concept of
“one” comes relatively quickly, but the concept of “two” frequently
includes anything “more than one” (e.g., •• or ••• are “two” or “not
one”; Carey & Sarnecka, 2006; Sarnecka & Lee, 2009; Wynn, 1990,
1992). Over time, children narrow their schema of “two” until it is
regularly applied to the quantity “two” (e.g., ••).
1.2. Mathematics-specific language
General language skills – particularly vocabulary – often
include words typically heard and used in everyday conversa-
tions and activities. However, individual content areas, especially
numeracy, have high content-specific vocabulary (Harmon et al.,
2005)—words that are primarily used in domain-related contexts
or have specific meanings in those contexts. Specifically, there are
two general areas of mathematical language terms that are often
discussed in relation to early numeracy development: quantita-
tive (Barner, Chow, & Yang, 2009) and spatial (Ramani, Zippert,
Schweitzer, & Pan, 2014). Understanding quantitative language
(such as “more,” “less,” “many,” and “fewer”) allows children
to make and describe comparisons between groups or numbers.
Understanding spatial language (such as “before,” “above,” and
“near”) allows children to talk about relations between physical
objects and between numbers on a number line. Essentially, much
of the terminology children need to understand and to “do” mathe-
matics is highly language-based. When children do not understand
aspects of early mathematical language such as quantifiers (e.g., all,
most, some) they also exhibit difficulties in grasping exact math-
ematical concepts such as cardinal number knowledge (Barner
et al., 2009). Further, in elementary school children, mathemati-
cal language is a strong predictor of numeracy skills even when
accounting for a range of cognitive covariates and prior numer-
acy achievement (Toll & Van Luit, 2014a). Moreover, knowledge
of mathematics-specific language has been surmised to be critical
toward the success of early mathematics curricula (Chard et al.,
2008; Clements & Sarama, 2011). Even the National Council of
Teachers of Mathematics (NCTM, 2006) indicates the importance
of high quality mathematics language in teaching:
“Children need introductions to the language and conventions of
mathematics, at the same time maintaining a connection to their
informal knowledge and language. They should hear mathemati-
cal language being used in meaningful contexts. . .Young children
need to learn words for comparing and for indicating position and
direction at the same time they are developing an understanding
of counting and number words” (NCTM, 2006).
Yet, even with the growing body of evidence supporting the
importance of mathematical language for numeracy development,
it is rarely a focus in early numeracy research. For example, in
the LeFevre et al. (2010) pathways model – as is the case in other
research connecting numeracy and language skills (Aunio et al.,
2006; Krajewski & Schneider, 2009; Purpura et al., 2011) – it is
presumed that the linguistic skills directly related to numeracy
development are general in nature – measured by components
of literacy and language such as vocabulary and/or phonological
awareness. Each of those studies included only a measure of gen-
eral language—and not a mathematical language measure. Further,
despite the consistent relation of general language skills to numer-
acy performance, improving children’s general language does not
actually result in positive impacts on children’s numeracy out-
comes (Jordan, Glutting, Dyson, Hassinger-Das, & Irwin, 2012).
One of the potential explanations for this finding is that general
language skills are acting as a proxy measure for another, more
proximal language skill—namely, mathematics-specific language.
1.3. Group differences in mathematical language
An abundance of evidence has indicated that children from fam-
ilies of low-SES perform significantly lower on numeracy (Jordan
et al., 1994; Starkey et al., 2004) and language (Hart & Risely,
1995) assessments than their peers from families of middle- and
higher-SES. The group differences in numeracy ability are pre-
sumed to be, at least in part, due to difficulties in language because
of the high language content of early numeracy skills (Jordan et al.,
1994; Jordan & Levine, 2009; Starkey et al., 2004). It is estimated
that by the time children from families with lower SES enter
kindergarten, they have heard, on average, approximately 20–30
million fewer words than their peers from families of higher-SES
(Hart & Risely, 1995). Some of the early language in which chil-
dren and their families engage is mathematical language (Durkin,
Shire, Riem, Crowther, & Rutter, 1986; Fuson, 1988). Unfortunately,
parents from lower-SES engage in fewer and less complex math-
ematical interactions with their children than their counterparts
from middle- and higher-SES (Saxe, Guberman, & Gearheart, 1987;
Vandermaas-Peeler, Nelson, Bumpass, & Sassine, 2009). Thus, chil-
dren from families of lower-SES are, from very early on, less likely to
have had the opportunity to acquire mathematical language skills;
yet, no study to date has investigated group differences in children’s
knowledge of mathematical language words.
 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268
1.4. Current study
Despite the agreement that mathematical language is impor-
tant for numeracy skills and that it begins to be acquired relatively
early on in a child’s development, little attention has been devoted
toward understanding and evaluating individual and group differ-
ences in children’s knowledge of mathematical language. In fact,
most studies evaluating the relation between language and early
numeracy skills have focused on general measures of language
and/or literacy (Aunio et al., 2006; Krajewski & Schneider, 2009;
LeFevre et al., 2010; Purpura et al., 2011). Of the few studies that
have examined children’s knowledge of mathematical language,
none has evaluated group differences. Thus, there are two broad
research questions guiding this study.
Research Question 1: Do individual differences in mathe-
matical language predict numeracy skills above and beyond
individual differences in general language? The primary purpose
of this study was to determine if the variance accounted for by
general language skills in predicting numeracy performance was
better accounted for by children’s knowledge of mathematical lan-
guage. It was hypothesized that mathematical language, and not
general language, would be a unique predictor of numeracy skills
because targeted content language is necessary for understanding
and completing numeracy tasks. To adequately address this issue,
two specific goals must be met:
1. General language must be a significant predictor of early numer-
acy skills when controlling for background variables.
2. Mathematical language must be a significant predictor of early
numeracy skills above and beyond general language skills when
accounting for the same background variables.
Research Question 2: Are there group differences in math-
ematical language performance based on parental education
and age? A secondary goal of this study was to explore group
differences in mathematical language in relation to parental edu-
cation levels (using parental education as a proxy for SES) and
age. Similar to previous findings on general numeracy and lan-
guage tasks, it was hypothesized that children from families with
low parental education would perform significantly lower on the
measure of mathematical language compared to their peers from
families with middle- and high-parental education. Further, as
would be expected, it was hypothesized that older children will
have higher levels of mathematical language than will younger chil-
dren, but that even the 3-year-old children will have developed a
sizeable base of mathematical language knowledge.
2. Method
2.1. Participants
Data were collected in 12 private preschools/kindergartens
serving children from families of a range of socioeconomic sta-
tuses. Parents of 147 children completed the consent forms for
the study. Of these children, four left their school before testing
was completed and seven children did not assent to participate.
The 136 children who completed the assessments were approxi-
mately evenly split by gender (54.4% female) and representative
of the demographics of the area (69.1% Caucasian, 8.8% Asian, 2.9%
African–American, 2.9% Hispanic, 16.3% other/multi-racial). Chil-
dren ranged in age from 3.12 years to 5.94 years (M = 4.28 years,
SD = .67 years), were English speaking as noted by teachers and
assessors, and had no known developmental disorders. With regard
to parental education, 25.0% of participants had parents with less
than a college degree, 33.8% had at least one parent with a college
261
degree but no graduate degree, and 41.2% had at least one par-
ent with a postgraduate degree. This work was approved by the
Institutional Review Board.
2.2. Measures
2.2.1. Early numeracy
The Preschool Early Numeracy Skills Test—Brief Version (PENS-B)
is a 24-item numeracy task (˛ = .93). The 24 items are repre-
sentative of the broad numeracy skills children are expected to
acquire in preschool and kindergarten. Specific areas of assess-
ment include: set comparison, numeral comparison, one-to-one
correspondence, number order, identifying numerals, ordinal-
ity, and number combinations. The PENS-B takes approximately
five minutes to administer and children answered all 24 items.
They received one point for each correct answer. However, an
empirically-derived ceiling rule was applied where, after a child
incorrectly responded to three items in a row, they did not receive
points for items after the ceiling (results were comparable regard-
less of the use of the ceiling rule, but the rule was utilized to be
consistent with the developed standards for the measure; Purpura,
Reid, Eiland, & Baroody, 2015). The test has been shown to have
evidence of reliability using both classical test theory and item
response theory metrics (Purpura et al.). Furthermore, the measure
has evidence of convergent validity as it is highly correlated with
the Test of Early Mathematics Ability – 3rd Edition (r = .73; Ginsburg
& Baroody, 2003) and evidence of discriminant validity as it was
found to be more related to another measure of mathematics than
to a measure of early literacy (Purpura et al.).
2.2.2. General language
The Definitional Vocabulary subtest of the Test of Preschool Early
Literacy Skills (TOPEL; Lonigan, Wagner, Torgesen, & Rashotte, 2007)
is a measure of children’s depth of vocabulary. Given its high
correlations with other general language tasks (e.g., r = .71 with
the Expressive One-Word Picture Vocabulary Test; Lonigan et al.,
2007) and the apparent unidimensional structure of oral language
skills in young children (Foorman, Herrera, Petscher, Mitchell, &
Truckenmiller, 2015), the Definitional Vocabulary subtest was used
as a proxy for general language skills. Definitional vocabulary mea-
sures children’s single-word spoken vocabulary and their ability
to formulate definitions for words. Children are asked to identify
images and then asked to describe its purpose or use. The internal
consistency for this task is .94 for 3- to 5-year-old children.
2.2.3. Mathematical language
The mathematical language test is an author-developed mea-
sure of mathematical content language. In contrast to the PENS-B (a
measure of numeracy knowledge) which assessed children’s ability
to work with exact numbers (e.g., count, compare, add), the math-
ematical language measure assessed children’s understanding of
the key words used in early mathematics (e.g., a lot, more, nearest,
far away). Children were initially assessed on a battery of 33 quan-
titative and spatial mathematical language items that were each
administered once in a fixed order. The quantitative language items
included both expressive and receptive items and the spatial lan-
guage items were all receptive. The items were originally selected
by reviewing both state and national standards/foundations for
mathematical language terms commonly used in preschool instruc-
tion and assessment. Items that targeted these words were then
created by the first author and reviewed by three mathematics
education experts who were independent of the project.
All items were designed to be completed without exact quanti-
tative skills or in a non-numeracy context. For example, one of the
quantitative questions was designed to target a child’s understand-
ing of the term “more.” This item had two parts that needed to be
262 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268

responded to correctly for the child to receive credit. Children were Table 1
Difficulty and discrimination parameters for mathematical language items.
asked to compare two different sets of dots (e.g., 10 vs. 2 and 9 vs.
1) with such a gross difference that, regardless of a child’s numer- Key word(s) Language Type Discrimination Difficulty
acy ability, if they knew the meaning of the language term, they Take awaya Quantitative 0.71 −1.36
would be able to respond correctly. Similarly, an example of one of Nearest Spatial 0.71 −1.18
the spatial items was the item that targeted a child’s understanding Under Spatial 0.88 −1.10
of the term “nearest.” For this item, children were presented with A little bit Quantitative 0.68 −1.00
First (in line) Spatial 0.80 −0.94
a picture of a boy in the middle of the page and four dots spread
Far away Spatial 0.84 −0.82
around the page. One dot was immediately next to the boy and Most Quantitative 0.49 −0.77
the others were close to different edges of the page. Children were More Quantitative 0.72 −0.69
asked, “Point to the dot that is nearest to the boy.” Below Spatial 0.72 −0.64
Front Spatial 0.73 −0.59
Utilizing an item response theory framework (Embretson &
Middle Spatial 0.70 −0.48
Reise, 2000), the original 33 items were pared down to 16 final End Spatial 0.79 −0.22
items by removing those items that overlapped with other items or Last Spatial 0.68 −0.18
functioned poorly. IRT allows researchers to select items based on Fewest Quantitative 0.66 0.02
item-level characteristics. The item-level characteristics, or param- Before Spatial 0.62 0.13
Less Quantitative 0.64 0.32
eters, in a 2-PL model that describe item functioning are referred
a
to as the difficulty parameter and the discrimination parameter. Indicates an expressive item that used manipulatives. All other items were
receptive and used the testing manual with pictures.
The difficulty parameter measures the point along the ability spec-
trum at which a specific response option would be endorsed 50%
of the time for an individual with a given ability. Items with high (ICC = .15), and vocabulary (ICC = .01) measures. Three fixed-effect
difficulty parameters require a greater amount of latent ability covariates were also used in the analyses for the first research
to be answered correctly; hence, the item is more difficult to question. These covariates were age, gender, and highest parental
answer correctly than an item with a lower difficulty parameter. education of either parent (scored on an 8-point scale ranging from
The discrimination parameter measures how well an item differ- <8th grade to doctoral/postgraduate degree).
entiates between individuals with latent abilities above and below
the item’s difficulty parameter. If multiple items had compara- 2.3. Procedures
ble difficulty parameters, the one with the highest discrimination
parameter was retained for the final measure. 2.3.1. Assessment procedure
The final 16 selected items had a range of difficulty parameters Children were assessed on all tasks in the fall of the academic
(spanning the ability range) and strong discrimination parameters year. Assessments took place in the local preschools at times iden-
(indicating that the item could be used to effectively differentiate tified by the schools in a room designated by the individual school
individuals above and below a given latent ability). The difficulty directors or teachers. The order of task administration was counter-
and discrimination parameters of each of the final 16 items are balanced. Individuals who had either completed or were working
presented in Table 1. One quantitative item was expressive and toward completion of a Bachelor’s degree in psychology or human
required manipulatives. All other items were receptive and used a development conducted the assessments. All testers completed
testing manual with pictures. This task had an internal consistency two 2–3 h training sessions, followed by individual practice, and a
of .85. testing-out session to ensure they were fluent with the assessment
measures and general principles of working with young children.
2.2.4. Rapid automatized naming Lead project staff evaluated each tester’s ability to adhere to the
Rapid automatized naming (RAN) was used as a general mea- language used in the testing manual, their knowledge of specific
sure of cognitive ability as it underlies both numeracy and literacy testing and scoring procedures, and their fluent knowledge of task
(Georgiou, Tziraki, Manolitsis, & Fella, 2013) and recent evidence administration.
has indicated a significant amount of variance in mathematical Research Question 1: Do individual differences in mathe-
ability is best explained by basic cognitive skills (Bailey, Watts, matical language predict numeracy skills above and beyond
Littlefield, & Geary, 2014). It was assessed through a picture naming individual differences in general language? To evaluate the
task and a color naming task—these two tasks were comparable to relation between language and numeracy skills, mixed-effects
the tasks used in the Georgiou et al. study and other standard mea- regression analyses were conducted (Raudenbush & Bryk, 2002).
sures of RAN. Specifically, children were initially asked to name In each of these analyses, school was included as a random effect
four common pictures (e.g., house, cat, car, pig). They were then to account for variance across schools. Age, gender, parental edu-
presented with a page of 40 pictures (5 × 8) and asked to name the cation, and RAN were included as fixed-effect covariates. Age and
pictures in order as fast as they could. If a child incorrectly named parental education were both included as continuous variables. In
a picture or skipped a picture, they were redirected back to that Step 1, the analyses were conducted with just the covariates. In
picture. The total time (in seconds) it took to name all 40 pictures Step 2a, the general language (vocabulary) measure was added to
was recorded. After the picture task, the same task was repeated, show that general language does have a significant relation to early
but with colors (blue, red, green, black). The two measures of RAN numeracy skills. In Step 2b, the mathematical language measure
were highly correlated (r = .67). The average of the two completion was included in the model instead of the general vocabulary mea-
times was used as the child’s RAN score. sure to show that it was independently related to early numeracy
skills. Finally, in Step 3, the mathematical language measure and
2.2.5. Covariates the general language measure were added to the model together
School was included in the model as a random-effect covariate to determine if mathematical language was a significant predictor
to control for school-level effects for the first research question. above and beyond general language. The amount of unique vari-
There were 12 schools with an average of 11.3 students per school. ance explained by individual predictors is presented as pseudo-sr2
Only minimal variance in the academic variables can be attributed (pseudo-squared semipartial).
to schools as the unconditional intra-class correlation coefficients Research Question 2: Are there group differences in math-
(ICC) were low for the PENS-B (ICC = .17), mathematical language ematical language performance based on parental education
 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268
and age? To evaluate differences in mathematical language based
on age, children were divided into three age groups 3-year-olds
(n = 47), 4-year-olds (n = 68), and 5-year-olds (n = 21). To evaluate
individual differences in mathematical language based on parental
education, children were divided into three groups based on the
parent education variable: no college degree (n = 34), 2- or 4-year
college degree (n = 46), or postgraduate degree (n = 56). ANOVAs
with pairwise comparisons were conducted to evaluate group dif-
ferences for age and parental education and effect size differences
(Cohen’s d) were used for group comparisons.
3. Results
3.1. Preliminary analyses
Means, standard deviations, skew, and kurtosis for all measures
are presented in Table 2. All tasks were normally distributed. In
terms of mathematical language knowledge, although on aver-
age children correctly responded to over 70% of the items, there
were significant individual differences in performance. Corre-
lations between tasks are presented in Table 3. Because the
correlations between the tasks were all significant, a test of corre-
lated correlations (Meng, Rosenthal, & Rubin, 1992) was conducted
to ensure that the mathematical language measure was not just
acting as another measure of numeracy or general language skills.
The correlation between the PENS-B and the mathematical lan-
guage measure was not significantly higher than the correlation
between the vocabulary measure and the mathematical language
measure (z = 0.73, p = .233) suggesting that mathematical language
was similarly related to both numeracy and general language. How-
ever, the correlation between the mathematical language measure
and the PENS-B and the correlation between the mathematical
language measure and the general language measure were both
significantly higher than the correlation between the PENS-B and
the general language measure (z = 2.69, p = .004; z = 1.99, p = .023).
Essentially, the mathematical language measure appears to assess
the commonality between domains and is not simply assessing
mathematics or language.
3.2. Primary analyses
3.2.1. Research Question 1: Do individual differences in
mathematical language predict numeracy skills above and beyond
individual differences in general language?
Conditional ICCs were <.01 for all analyses which indicated that
nearly no variance was attributable at the school level once covari-
ates were included in the models. This finding was expected as
Table 2
Means, standard deviations, range, skewness, and kurtosis of the sum scores of the numeracy, language, mathematical language and rapid automatized naming tasks.
Task M SD
Preschool Early Numeracy Skills Test – Brief 10.66
Definitional Vocabulary 51.37
Mathematical Language 11.49
Rapid Automatized Naming 79.17
Table 3
Correlations between the raw total scores of all tasks.
1. Preschool Early Numeracy Skills Test – Brief
2. Definitional Vocabulary
3. Mathematical Language
4. Rapid Automatized Naming
Note. All correlations were significant at p < .001. Correlations on the bottom are zero-order correlations and correlations on the top are partial correlations accounting for
age.
263
data were collected in the fall of the academic year and both age
and parental education were accounted for in the analyses. Results
of the mixed-effect regression analyses are presented in Table 4.
In Step 1, all four covariates were significant predictors of early
numeracy skills. In Step 2a, when general language was added to
the model, it was a significant predictor of early numeracy skills
(estimate = 0.07, p = .045, 95% CI [0.00, 0.13]) and it accounted for
2% unique variance in numeracy performance. When the mathe-
matical language measure was included in step 2b instead of the
general language measure, mathematical language was a signif-
icant predictor of early numeracy performance (estimate = 0.45,
p < .001, 95% CI [0.23, 0.68]) and accounted for 6% unique variance
in early numeracy performance. However, when the mathematical
language measure and the general language measure were included
in the same model in Step 3, mathematical language was a signifi-
cant predictor of early numeracy skills (estimate = 0.42, p = 001, 95%
CI [0.18, 0.67]) accounting for 4% unique variance in early numer-
acy performance, but the general language measure was no longer
a significant predictor (estimate = 0.03, p = .466, 95% CI [−0.04, to
0.09]).
3.2.2. Research Question 2: Are there group differences in
mathematical language performance based on parental education
and age?
3.2.2.1. Age-related differences. Using an ANOVA, there was a sig-
nificant difference in mathematical language based on age group,
F(2,133) = 19.67, p < .001. Pairwise comparisons revealed that there
were significant differences in mean performance across all three
age groups (p’s < .05; see Table 5). Furthermore, to ensure that the
mathematical language measure had adequate psychometric prop-
erties within each age-group, internal consistency, skewness, and
kurtosis we assessed for each age group and found to be acceptable
(Table 5).
3.2.2.2. Parental education-related differences. Using an ANOVA,
there was a significance difference in mathematical language based
on parental education, F(2, 133) = 8.73, p < .001. Further, the three
parental education groups did not differ in age F(2, 133) = 1.66,
p = .195, indicating that the performance differences were unlikely
a function of age differences. Children whose parents had no college
degree performed significantly lower on the mathematics language
measure than children whose parents with a 2- or 4-year col-
lege degree, t(78) = 3.62, p = .001 (95% CI [1.35, 4.66]), and children
whose parents had a postgraduate degree t(88) = 3.65, p < .001 (95%
CI [1.37, 4.63]). However, there was no significant difference in
performance on the mathematical language measure between chil-
dren whose parents had a 2- or 4-year degree and children whose
Range Skew Kurtosis
6.12 0 to 24 0.24 −0.89
12.58 8 to 69 −1.31 1.39
3.83 2 to 16 −0.82 −0.19
30.32 25.27 to 186.89 1.37 2.05
1 2 3 4
– .32 .53 −.45
.56 – .47 −.32
.69 .62 – −.44
−.59 −.48 −.57 –
264 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268

Table 4
Mixed effects regression predicting early numeracy performance.

Pseudo-R2 Estimate SE F p 95% CI Pseudo-sr2

Step 1 .45
Age 4.87 0.56 75.22 <.001 3.76 to 5.99 .29
Gender −1.64 0.68 5.78 .017 −2.98 to −0.29 .02
Parental Education 1.13 0.21 28.12 <.001 0.71 to 1.55 .05
Rapid Automatized Naming −0.06 0.01 21.22 <.001 −0.08 to −0.03 .08

Step 2a .47
Age 4.29 0.62 47.00 <.001 3.05 to 5.53 .19
Gender −1.46 0.68 4.67 .032 −2.80 to −0.12 .02
Parental Education 1.09 0.21 26.51 <.001 0.67 to 1.51 .04
Rapid Automatized Naming −0.05 0.01 14.29 <.001 −0.08 to −0.03 .06
Definitional Vocabulary 0.07 0.03 4.10 .045 0.00 to 0.13 .02

Step 2b .51
Age 3.81 0.60 40.53 <.001 2.62 to 4.99 .15
Gender −1.25 0.65 3.66 .058 −2.54 to 0.04 .01
Parental Education 0.84 0.21 15.42 <.001 0.42 to 1.27 .04
Rapid Automatized Naming −0.04 0.01 9.63 .002 −0.07 to −0.01 .03
Mathematical Language 0.45 0.11 15.63 <.001 0.23 to 0.68 .06

Step 3 .51
Age 3.66 0.63 33.95 <.001 2.42 to 4.91 .13
Gender −1.21 0.65 3.42 .067 −2.51 to 0.09 .01
Parental Education 0.85 0.21 15.54 <.001 0.42 to 1.27 .04
Rapid Automatized Naming −0.04 0.01 8.72 .004 −0.06 to −0.01 .03
Definitional Vocabulary 0.03 0.03 0.53 .466 −0.04 to 0.09 .00
Mathematical Language 0.42 0.12 11.68 .001 0.18 to 0.67 .04

N = 136.

Table 5
Descriptive statistics for mathematical language by age and parental education.

Age Group n Range M SD Skewness Kurtosis Effect Size

3-Year-Olds 47 2 to 16 9.15 4.02 −0.09 −1.07 0.87***

4-Year-Olds 68 2 to 16 12.26 3.17 −1.30 1.76 0.64*
5-Year-Olds 21 8 to 16 14.19 2.36 −1.29 0.94 –

Parental Education
No College Degree 34 2 to 16 9.24 4.11 0.04 −0.97 a
0.81** /0.79***
2- or 4-year Degree 46 3 to 16 12.24 3.31 −1.05 0.77 0.00ns
Postgraduate Degree 56 2 to 16 12.23 3.56 −1.32 1.38 –
a
The first effect size is for the difference between No College Degree and 2- or 4-year Degree and the second effect size is for the difference between No College Degree
and Postgraduate Degree.
*
p < .05.
**
p < .01.
***
p < .001.
ns
p > .05.

parents had a postgraduate degree, t(100) = 0.01, p = .992 (95% CI
[−1.36, 1.37]).
4. Discussion
Consistent with previous studies (Aunio et al., 2006; Krajewski
& Schneider, 2009; LeFevre et al., 2010; Purpura et al., 2011), the
findings from this study indicate that general language was a signif-
icant predictor of numeracy skills even when including a number
of key covariates. However, in extending beyond prior literature,
mathematical language was a significant predictor of numeracy
skills over-and-above general language in young children. Impor-
tantly, when mathematical language was included in the model,
general language was no longer a significant predictor of numer-
acy skills. Essentially, in prior studies where general language skills
were found to be significantly related to numeracy skills, it may
have actually been acting as a proxy measure for mathematical
language which is more proximal to numeracy performance (Toll
& Van Luit, 2014b). Notably, the group difference analyses indi-
cate that similar parental education-related differences were found
for mathematical language as have been found for both numeracy
skills and general language (Jordan et al., 1994; Jordan & Levine,
2009; Starkey et al., 2004). Further, it is evident that children begin
to acquire mathematical language early in development—as chil-
dren appear to have acquired a substantial body of mathematical
language even as early as three years old. This study provides evi-
dence for the importance of assessing mathematical language when
evaluating young children’s early numeracy skills.
4.1. Do individual differences in mathematical language predict
numeracy skills above and beyond individual differences in
general language?
Practitioners and researchers have been calling attention to
the importance of specific mathematical language for decades
across a large range of ages (Adams, 2003; Knight & Hargis, 1977;
Lansdell, 1999; Schleppegrell, 2007). Recently, this call has been
echoed in major early mathematics policy documents as well
(Clements, Baroody, & Sarama, 2014; NCTM, 2006). However, very
little research has examined the predictive value of mathematics-
specific language on early numeracy performance in young children
(Toll & Van Luit, 2014a; 2014b). Though other work has identified
relations between children’s production of math words and their
numeracy performance in preschool (Jennings, Jennings, Richey, &
 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268
Dixon-Krauss, 1992), such studies only account for the amount of
words used (even if the same word is used repeatedly) and do not
account for the breadth of mathematical language. This is the first
study to investigate individual and group differences in breadth of
mathematical language with children as young as 3- and 4-years
of age. Though prior research (Toll & Van Luit, 2014a; 2014b) did
include some 4-year-olds in their studies, mathematical language
was not assessed until children were, on average, 5.5-years old.
It is evident through the current findings, and through previous
research (Barner et al., 2009; Durkin et al., 1986; Fuson, 1988;
Sarnecka et al., 2007), that children acquire these skills early on
in their development.
Most prior work investigating the link between mathematical
development and language has utilized measures of general lan-
guage skills which, ultimately, may have been acting as a proxy
measure for mathematical language. Those that have evaluated
children’s mathematical language have utilized measures of total
production (Jennings et al., 1992) or measures that were not orig-
inally intended to measure mathematical language (Toll & van
Luit, 2014a, 2014b). The measure utilized in the present study
was designed to intentionally include a range of both quantita-
tive and spatial mathematical language and specifically to measure
children’s breadth of knowledge of these words. As children’s
knowledge of mathematical language is a more proximal measure
to numeracy skills than are general language skills, more consider-
ation needs to be given to the assessment of children’s knowledge
of mathematical language in future studies.
It is important to note that the even though the total variance
explained in this model was relatively high (51%), the amount of
unique variance in numeracy skills accounted for by mathematical
language was small (4%) according to criteria established by Cohen
(1992). Yet, aside from age, mathematical language accounted for
more unique variance than the other variables included in the
model. These results align with recent findings by Bailey et al.
(2014) who found that a large portion of the variance in predicting
numeracy skills is attributable to general cognitive and background
factors—much of which is shared variance. Given that mathemati-
cal language accounts for unique variance above and beyond these
strong covariates, and it is more likely to be malleable than general
cognitive and background factors, it may be a potential domain for
targeted intervention.
4.2. Are there group differences in mathematical language
performance based on parental education and age?
This study is the first to report on group differences in young
children’s knowledge of mathematical language. The results of this
study indicate that not only are there significant differences in
performance on mathematical language amongst 3-, 4-, and 5-year-
olds, as one might expect, but that even the youngest children had
a sizable understanding of mathematical vocabulary. Critically, it
was also found that children from families with at least one college-
educated parent performed better on the measure of mathematical
language than children whose parents had no college degree. A
possible explanation for this variation may be that parents with
higher levels of education more often provide richer home numer-
acy environments for their children. Caregiver-child interactions
during informal activities play an important role in children’s cog-
nitive development (Rogoff, 1990; Vgotsky, 1978), including in
their mathematical thinking (Blevins-Knabe & Musun-Miller, 1996;
Vandermaas-Peeler, 2008). Parents with higher levels of education
often hold higher expectations for their children’s mathematical
skill attainment, and have a better sense of which mathematical
skills are within their children’s developmental range (DeFlorio &
Beliakoff, 2015). Thus, they are likely able to provide more direct,
265
frequent, and appropriate mathematical experiences for their chil-
dren at home.
In addition to providing opportunities for their children to
engage in numeracy activities, parents’ own math talk is associ-
ated with children’s numeracy development (Gunderson & Levine,
2011; Levine, Suriyakham, Rowe, Huttenlocher, & Gunderson,
2010; Ramani, Rowe, Eason, & Leech, 2015). Given the signifi-
cant group differences in home language environment in general
(Hart & Risley, 1995), these differences may extend to mathemat-
ical language as well. For example, Ramani et al. (2015) found
that during parent–child interactions, parental foundational math
utterances (e.g., counting, number identification) and advanced
concepts math talk (e.g., cardinality, ordinal relations, and arith-
metic) were positively and significantly associated with children’s
math foundational and advanced math utterances. However, it is
important to note that not all math talk is not equally effective
(Boonen, Kolkman, & Kroesbergen, 2011; Lansdell, 1999) because
it requires a level of content knowledge and pedagogical skill that
many parents and some early childhood educators may not have
had the opportunity to acquire (Ginsburg, Lee, & Boyd, 2008). Crit-
ically, it appears that it is not just the quantity of early language
input that is important for children’s development, but it is also
the quality of such input (Hirsh-Pasek et al., 2015).
Given the proximal nature of mathematical language to numer-
acy skills and the finding of parental education-based performance
differences in mathematical language, children from families with
low parental education may be those who can benefit the most
from home and school environments rich in mathematical lan-
guage. There is a need to better understand why these differences
exist and how to prevent or remediate these differences. By iden-
tifying methods to provide support and guidance to parents and
teachers on how to enrich a child’s environment with mathematical
language – particularly the home and school environments of chil-
dren from families with lower parental education – they may have
more success in promoting numeracy development in the young
children who are most at-risk of later mathematical difficulties.
4.3. Limitations and future directions
Several limitations to the findings in this study and subsequent
future directions for research must be discussed. First, though
these findings provide evidence that suggests mathematical lan-
guage is predictive of early numeracy skills; these findings are
only correlational in nature. Future causal research on understand-
ing the mechanism(s) by which mathematical language affects
numeracy skill development need to be investigated as well as the
identification of other factors that may support this development
such as parent and teacher mathematical language interactions.
Although Hassinger-Das, Jordan, and Dyson (2015) have conducted
an intervention with kindergarten children on mathematical lan-
guage, they did not find direct transfer to general numeracy
knowledge. They did, however, provide evidence that mathemati-
cal language is malleable through intervention. Importantly, their
study was not designed to assess the causal mechanism of mathe-
matical language influencing numeracy development because their
instruction involved both mathematical language and numeracy
content. Specifically, in the mathematical language instructional
condition, children first engaged in storybook reading and direct
instruction related to specific vocabulary concepts and then they
practiced the related numeracy skills. The immediate connection
between the mathematical language concepts and the numeracy
skills may not have provided sufficient time for the children to fully
understand the language terms before applying them to numeracy
skills. Thus, the ordering and connection of mathematical lan-
guage and numeracy instruction (prior to or in conjunction with)
should be investigated to determine if mathematical language is a
266 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268
necessary precursor to acquisition of numeracy skills or if it simply
enhances the acquisition of concurrent instruction.
Second, and related to the first limitation, these findings are
concurrent and thus, the directionality of the relation between
numeracy and language cannot be fully ascertained. Though some
studies have shown general mathematics skills to be a strong pre-
dictor of later literacy skills (Claessens, Duncan, & Engel, 2009;
Claessens & Engel, 2013; Duncan et al., 2007; Hecht, Torgesen,
Wagner, & Rashotte, 2001; Hooper et al., 2010), the rationale for
why it is such as strong predictor is not well understood. Claessen
and Engel (2013) posit that the relation may be the result of the
complex features of early mathematics skills aligning with the com-
plex features of later language. Given the significant amount of
unique variance accounted for in early numeracy by mathemat-
ical language, the language component of mathematics may be
one of those complex aspects accounting for significant variance
in later literacy skills. As such, further investigation of both how
language-related skills influence mathematics development and
how mathematics skills may influence language development are
needed.
Third, the constructs assessed in this study were relatively
restricted and may not fully represent (a) the broader domains
of mathematics and language or (b) the subcomponents of each
domain. Numeracy and vocabulary are both only aspects of their
broader constructs (mathematics and oral language). Mathematics
broadly includes numeracy, geometry, measurement, and pattern-
ing. Similarly, oral language skills include vocabulary, syntactical
awareness, and comprehension. Though these aspects of math-
ematics and language are highly-related within domains, some
may involve mathematical language more than others. Further-
more, within the domains of numeracy and vocabulary, there are
multiple subcomponents such as counting, relational, and opera-
tional skills for numeracy, and both expressive and receptive skills
for vocabulary. Different aspects of numeracy (e.g., set compari-
son, cardinality, story problems) may be differentially connected
to aspects of mathematical language (e.g., quantitative and spa-
tial) or language skills in general. There are also other, non-verbal
mathematics skills such as the approximate number sense and
approximate addition that may show no relation to language
(Jordan, Kaplan, Oláh, & Locuniak, 2006). Additionally, though lan-
guage skills are highly related to one another in young children, the
use of receptive as well as expressive measures may have yielded
a more nuanced understanding of the relation between language
and mathematics. At present, these findings cannot be extended to
the broader or more targeted constructs in mathematics and lan-
guage. Future research that includes a depth of both mathematics
and language components should be conducted to better delineate
the nuanced relations between domains.
Fourth, the sample was relatively homogenous and had higher
levels of parental education than the general population. These
findings may not be generalizable to more heterogeneous popula-
tions and specifically to populations predominantly from families
with lower SES because parents with higher levels of education
typically spend more time with their children (Guryan, Hurst, &
Kearney, 2008). Harding (2015) found that increases in parental
(specifically maternal) education are related to small, but reliable,
increases in children’s early mathematics and language skills. It
has been surmised that the early academic performance improve-
ments associated with mothers’ increases in education are due to
increases in the frequency of parent–child learning interactions
(Magnuson, Sexton, Davis-Kean, & Huston, 2009). Thus, it is possi-
ble that the relation between language – particularly mathematical
language – and numeracy skills differ as a function of SES because
of potential parental influences. Additional longitudinal research
with more diverse samples is needed.
5. Conclusion
The findings from the current study expand on the existing
literature-base by identifying not only that mathematical language
is a more proximal predictor of numeracy performance than is gen-
eral language, but also that it begins to develop relatively early on
and children from families with low parental education tend have
acquired significantly less mathematical language than their peers
from families with middle- and higher-parental education. This evi-
dence indicates that this understudied component of the linguistic
pathway may be critical for successful acquisition of early numer-
acy skills. Future work identifying the mechanism(s) by which
mathematical language affects numeracy skill development and
how best to enhance young children’s mathematical language skills
are needed. Ultimately, given that these skills appear to develop
quite early – even by the time a child is 3-years old – developing
effective methods for both parents and early childhood education
teachers to provide mathematical language rich environments may
be warranted.
References
Adams, T. L. (2003). Reading mathematics: more than words can say. The Reading
Teacher, 56, 786–795.
Aunio, P., Niemivirta, M., Hautamaki, J., Van Luit, J. E. H., Shi, J., & Zhang, M. (2006).
Young children’s number sense in China and Finland. Scandinavian Journal of
Educational Research, 50, 483–502. http://dx.doi.org/10.1080/
00313830600953576
Bailey, D. H., Watts, T. W., Littlefield, A. K., & Geary, D. C. (2014). State and trait
effects on individual differences in children’s mathematical development.
Psychological Science, 25, 2017–2026. http://dx.doi.org/10.1037/t15170-000
Baldo, J. V., & Dronkers, N. F. (2007). Neural correlates of arithmetic and language
comprehension: a common substrate? Neuropsychologia, 45, 229–235. http://
dx.doi.org/10.1016/j.neuropsychologia.2006.07.014
Barner, D., Chow, K., & Yang, S. (2009). Finding one’s meaning: a test of the relation
between quantifiers and integers in language development. Cognitive
Psychology, 58, 195–219. http://dx.doi.org/10.1016/j.cogpsych.2008.07.001
Blevins-Knabe, B., & Musun-Miller, L. (1996). Number use at home by children and
their parents and its relationship to early mathematical performance. Early
Development and Parenting, 5, 35–45. http://dx.doi.org/10.1002/(SICI)1099-
0917(199603)5:1<35::AID-EDP113>3.0.CO;2-0
Boonen, A. J. H., Kolkman, M. E., & Kroesbergen, E. H. (2011). The relation between
teachers’ math talk and the acquisition of number sense with kindergarten
classrooms. Journal of School Psychology, 49, 281–299. http://dx.doi.org/10.
1016/j.jsp.2011.03.002
Carey, S., & Sarnecka, B. W. (2006). The development of human conceptual
representations. In M. Johnson, & Y. Munakata (Eds.), Processes of change in
brain and cognitive development: attention and performance XXI (pp. 473–496).
Oxford, UK: Oxford University Press.
Chard, D. J., Baker, S. K., Clarke, B., Jungjohann, K., Davis, K., & Smolkowski, K.
(2008). Preventing early mathematics difficulties: the feasibility of a rigorous
kindergarten mathematics curriculum. Learning Disability Quarterly, 31, 11–20.
Claessens, A., Duncan, G., & Engel, M. (2009). Kindergarten skills and fifth-grade
achievement: evidence from the ECLS-K. Economics of Education Review, 28,
415–427. http://dx.doi.org/10.1016/j.econedurev.2008.09.003
Claessens, A., & Engel, M. (2013). How important is where you start? Early
mathematics knowledge and later school success. Teachers College Record, 115,
1–29.
Clements, D. H., & Sarama, J. (2011). Early childhood mathematics intervention.
Science, 333, 968–970. http://dx.doi.org/10.1126/science.1204537
Clements, D. H., Baroody, A. J., & Sarama, J. (2014). Background research on early
mathematics: background research for the National Governor’s Association (NGA)
center project on early mathematics. Washington, D.C: National Governor’s
Association. http://www.nga.org/files/live/sites/NGA/files/pdf/2013/
1311SEME-Background.pdf
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
DeFlorio, L., & Beliakoff, A. (2015). Socioeconomic status and preschoolers’
mathematical knowledge: the contribution of home activities and parent
beliefs. Early Education and Development, 26, 319–341. http://dx.doi.org/10.
1080/10409289.2015.968239
Dehaene, S. (1992). Varieties of numerical abilities. Cognition, 44, 1–44. http://dx.
doi.org/10.1016/0010-0277(92)90049-N
Dehaene, S. (1997). The number sense: how the mind creates mathematics. New York,
NY: Oxford University Press.
Dehaene, S., Spelke, E., Pinel, P., Stanescu, R., & Tsivkin, S. (1999). Sources of
mathematical thinking: behavioral and brain-imaging evidence. Science, 284,
970–974. http://dx.doi.org/10.1126/science.284.5416.970
 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268
Duncan, G. J., Dowsett, C. J., Claessens, A., Magnuson, K., Huston, A. C., Klebanov, P.,
. . . & Japel, C. (2007). School readiness and later achievement. Developmental
Psychology, 43, 1428–1446. http://dx.doi.org/10.1037/0012-1649.43.6.1428
Durkin, K., Shire, B., Riem, R., Crowther, R. D., & Rutter, D. R. (1986). The social and
linguistic context of early number word use. British Journal of Developmental
Psychology, 4, 269–288. http://dx.doi.org/10.1111/j.2044-835X.1986.tb01018.x
Embretson, S. E., & Reise, S. P. (2000). Item response theory of psychologists.
Mahwah. NJ: Erlbaum.
Foorman, B. R., Herrera, S., Petscher, Y., Mitchell, A., & Truckenmiller, A. J. (2015).
The structure of oral language and its relation to reading at various levels of
comprehension in kindergarten through grade 2. Reading and Writing, http://
dx.doi.org/10.1007/s11145-015-9544-5
Fuson, K. C. (1988). Children’s counting and concepts of number. New York, NY:
Springer-Verlag.
Georgiou, G. K., Tziraki, N., Manolitsis, G., & Fella, A. (2013). Is rapid automatized
naming related to reading and mathematics for the same reason(s)? A
follow-up study from kindergarten to grade 1. Journal of Experimental Child
Psychology, 115, 481–496. http://dx.doi.org/10.1016/j.jecp.2013.01.004
Ginsburg, H. P., & Baroody, A. J. (2003). Test of early mathematics ability (3rd ed.).
Austin, TX: Pro-ed.
Ginsburg, H. P., Lee, J. S., & Boyd, J. S. (2008). Mathematics education for young
children: what it is and how to promote it. Social Policy Report, 22, 3–23.
Gordon, P. (2004). Numerical cognition without words: evidence from Amazonia.
Science, 306, 496–499. http://dx.doi.org/10.1126/science.1094492
Guryan, J., Hurst, E., & Kearney, M. (2008). Parental education and parental time
with children. Journal of Economic Perspectives, 22, 23–46. http://dx.doi.org/10.
1257/jep.22.3.23
Gunderson, E. A., & Levine, S. C. (2011). Some types of parent number talk count
more than others: relations between parents’ input and children’s
cardinal-number knowledge. Developmental Science, 14, 1021–1032. http://dx.
doi.org/10.1111/j.1467-7687.2011.01050.x
Halberda, J., Mazzocco, M., & Feigenson, L. (2008). Individual differences in
nonverbal number acuity predict maths achievement. Nature, 455, 665–668.
http://dx.doi.org/10.1038/nature07246
Hanich, L. B., Jordan, N. C., Kaplan, D., & Dick, J. (2001). Performance across
different areas of mathematical cognition in children with learning difficulties.
Journal of Educational Psychology, 93, 615–626. http://dx.doi.org/10.1037/0022-
0663.93.3.615
Harding, J. F. (2015). Increases in maternal education and low-income children’s
cognitive and behavioral outcomes. Developmental Psychology, 51, 583–599.
http://dx.doi.org/10.1037/a0038920
Harmon, J. M., Hedrick, W. B., & Wood, K. D. (2005). Research on vocabulary
instruction in the content areas: implications for struggling readers. Reading &
Writing Quarterly, 21, 261–280. http://dx.doi.org/10.1080/
10573560590949377
Hart, B., & Risley, T. R. (1995). Meaningful differences in the everyday experiences of
young American children. Baltimore, MD: Paul H. Brookes.
Hassinger-Das, B., Jordan, N. C., & Dyson, N. (2015). Reading stories to learn math:
mathematics vocabulary instruction for children with early numeracy
difficulties. Elementary School Journal, 116, 242–264. http://dx.doi.org/10.1086/
683986
Hecht, S. A., Torgesen, J. K., Wagner, R. K., & Rashotte, C. A. (2001). The relations
between phonological processing abilities and emerging individual differences
in mathematical computational skills: a longitudinal study from second to fifth
grade. Journal of Experimental Child Psychology, 79, 192–227. http://dx.doi.org/
10.1006/jecp.2000.2586
Hirsh-Pasek, K., Adamson, L. B., Bakeman, R., Owen, M. T., Golinkoff, R. M., . . . &
Suma, K. (2015). The contribution of early communication quality to
low-income children’s language success. Psychological Science, http://dx.doi.
org/10.1177/0956797615581493
Hooper, S. R., Roberts, J., Sideris, J., Burchinal, M., & Zeisel, S. (2010). Longitudinal
predictors of reading and math trajectories through middle school from
African American versus Caucasian students across two samples.
Developmental Psychology, 46, 1018–1029. http://dx.doi.org/10.1037/a0018877
Jennings, C. M., Jennings, J. E., Richey, J., & Dixon-Krauss, L. (1992). Increasing
interest and achievement in mathematics through children’s literature. Early
Childhood Research Quarterly, 7, 263–276. http://dx.doi.org/10.1016/0885-
2006(92)90008-M
Jordan, N. C., & Hanich, L. B. (2000). Mathematical thinking in second-grade
children with different forms of LD. Journal of Learning Disabilities, 33, 567–578.
http://dx.doi.org/10.1177/002221940003300605
Jordan, N. C., & Levine, S. C. (2009). Socioeconomic variation, number competence,
and mathematics learning difficulties in young children. Developmental
Disabilities Research Reviews, 15, 60–68. http://dx.doi.org/10.1002/ddrr.46
Jordan, N. C., Huttenlocher, J., & Levine, S. C. (1994). Assessing early arithmetic
abilities: effects of verbal and nonverbal response types on the calculation
performance of middle- and low-income children. Learning and Individual
Differences, 6, 413–432. http://dx.doi.org/10.1016/1041-6080(94)90003-5
Jordan, N. J., Glutting, J., Dyson, N., Hassinger-Das, B., & Irwin, C. (2012). Building
kindergartners’ number sense: a randomized controlled study. Journal of
Educational Psychology, 104, 647–660. http://dx.doi.org/10.1037/a0029018
Jordan, N. C., Kaplan, D., Oláh, L. N., & Locuniak, M. N. (2006). Number sense growth
in kindergarten: a longitudinal investigation of children at risk for
mathematics difficulties. Child Development, 77, 153–175. http://dx.doi.org/10.
1111/j.1467-8624.2006.00862.x
267
Juel, C. (1988). Learning to read and write: a longitudinal study of 54 children from
first through fourth grades. Journal of Educational Psychology, 80, 437–447.
http://dx.doi.org/10.1037/0022-0663.80.4.437
Knight, L. N., & Hargis, C. H. (1977). Math language ability: its relationship to
reading in math. Language Arts, 54, 423–428.
Krajewski, K., & Schneider, W. (2009). Early development of quantity to
number-word linkage as a precursor of mathematical school achievement and
mathematical difficulties: findings from a four-year longitudinal study.
Learning and Instruction, 19, 513–526. http://dx.doi.org/10.1016/j.learninstruc.
2008.10.002
Lansdell, J. M. (1999). Introducing young children to mathematical concepts:
Problems with ‘new’ terminology. Educational Studies, 25, 327–333. http://dx.
doi.org/10.1080/03055699997837
LeFevre, J.-A., Fast, L., Skwarchuk, S.-L., Smith-Chant, B. L., Bisanz, J., Kamawar, D., &
Penner-Wilger, M. (2010). Pathways to mathematics: longitudinal predictors
of performance. Child Development, 81, 1753–1767. http://dx.doi.org/10.1111/j.
1467-8624.2010.01508.x
Levine, S. C., Suriyakham, L. W., Rowe, M. L., Huttenlocher, J., & Gunderson, E. A.
(2010). What counts in the development of young children’s number
knowledge? Developmental Psychology, 46, 1309–1319. http://dx.doi.org/10.
1037/a0019671
Lewis, C., Hitch, G. J., & Walker, P. (1994). The prevalence of specific arithmetic
difficulties and specific reading difficulties in 9- to 10-year old boys and girls.
Child Psychology & Psychiatry & Allied Disciplines, 35, 283–292. http://dx.doi.
org/10.1111/j.1469-7610.1994.tb01162.x
Li, X., Sun, Y., Baroody, A. J., & Purpura, D. J. (2013). The effect of language on
Chinese and American 2- and 3-year olds’ small number identification.
European Journal of Psychology of Education, 28, 1525–1542. http://dx.doi.org/
10.1007/s10212-013-0180-7
Lonigan, C. J., Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (2007). Test of preschool
early literacy. Austin, TX: Pro-ed.
Magnuson, K., Sexton, H. R., Davis-Kean, P. E., & Huston, A. C. (2009). Increases in
maternal education and young children’s language skills. Merrill-Palmer
Quarterly, 55, 319–350. http://dx.doi.org/10.1353/mpq.0.0024
Mann Koepke, K., & Miller, B. (2013). At the intersection of math and reading
disability: introduction to the special issue. Journal of Learning Disabilities, 46,
483–489. http://dx.doi.org/10.1177/0022219413498200
McClelland, M. M., Acock, A. C., & Morrison, F. J. (2006). The impact of kindergarten
learning-related skills on academic trajectories at the end of elementary
school. Early Childhood Research Quarterly, 21, 471–490. http://dx.doi.org/10.
1016/j.ecresq.2006.09.003
Meng, X.-L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated correlation
coefficients. Psychological Bulletin, 111, 172–175.
Miura, I. T., & Okamoto, Y. (2003). Language supports for mathematics
understanding and performance. In A. J. Baroody, & A. Dowker (Eds.), The
development of arithmetic concepts and skills: constructing adaptive expertise
(pp. 229–242). Mahwah, NJ: Erlbaum.
Mix, K. S. (2009). How Spencer made number: first uses of the number words.
Journal of Experimental Child Psychology, 102, 427–444. http://dx.doi.org/10.
1016/j.jecp.2008.11.003
National Council of Teachers of Mathematics. (2006). Curriculum focal points for
prekindergarten through grade 8 mathematics. Reston, VA: NCTM.
Purpura, D. J., & Ganley, C. (2014). Working memory and language: Skill-specific or
domain-general relations to mathematics? Journal of Experimental Child
Psychology, 122, 104–121. http://dx.doi.org/10.1016/j.jecp.2013.12.009
Purpura, D. J., Hume, L., Sims, D., & Lonigan, C. J. (2011). Early literacy and early
numeracy: the value of including early literacy skills in the prediction of
numeracy development. Journal of Experimental Child Psychology, 110,
647–658. http://dx.doi.org/10.1016/j.jecp.2011.07.004
Purpura, D. J., Reid, E. E., Eiland, M. D., & Baroody, A. J. (2015). Using a brief
preschool early numeracy skills screener to identify young children with
mathematics difficulties. School Psychology Review, 44, 41–59. http://dx.doi.
org/10.17105/SPR44-1.41-59
Ramani, G. B., Rowe, M. L., Eason, S. H., & Leech, K. A. (2015). Math talk during
information learning activities in Head Start families. Cognitive Development,
35, 15–33. http://dx.doi.org/10.1016/j.cogdev.2014.11.002
Ramani, G. B., Zippert, E., Schweitzer, S., & Pan, S. (2014). Preschool children’s joint
block building during a guided play activity. Journal of Applied Developmental
Psychology, 35, 326–336. http://dx.doi.org/10.1016/j.appdev.2014.05.005
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: applications and
data analysis methods (2nd ed.). London, England: Sage.
Rogoff, B. (1990). Apprenticeship in thinking: cognitive development in social context.
New York, NY: Oxford University Press.
Romano, E., Babchishin, L., Pagani, L. S., & Kohen, D. (2010). School readiness and
later achievement: replication and extension using a nationwide Canadian
survey. Developmental Psychology, 46, 995–1007. http://dx.doi.org/10.1037/
a0018880
Sarama, J., & Clements, D. H. (2009). Early childhood mathematics education
research: learning trajectories for young children. New York, NY: Routledge.
Sarnecka, B. W., & Lee, M. D. (2009). Levels of number knowledge during early
childhood. Journal of Experimental Child Psychology, 103, 325–337. http://dx.
doi.org/10.1016/j.jecp.2009.02.007
Sarnecka, B. W., Kamenskaya, V. G., Yamana, Y., Ogura, T., & Yudovina, J. B. (2007).
From grammatical number to exact numbers: early meanings of one, ‘two,’ and
three in English, Russian, and Japanese. Cognitive Psychology, 55, 136–168.
http://dx.doi.org/10.1016/j.cogpsych.2006.09.001
268 D.J. Purpura, E.E. Reid / Early Childhood Research Quarterly 36 (2016) 259–268
Saxe, G. B., Guberman, S. R., & Gearhart, M. (1987). Social processes in early
number development. Monographs of the Society for Research in Child
Development, 52 i-162 http://dx.doi.org/10.2307/1166071
Schleppegrell, M. J. (2007). The linguistic challenges of mathematics teaching and
learning: a research review. Reading & Writing Quarterly, 23, 139–159. http://
dx.doi.org/10.1080/10573560601158461
Silver, C. H., Pennett, D.-L., Black, J. L., Fair, G. W., & Balise, R. R. (1999). Stability of
arithmetic disability subtypes. Journal of Learning Disabilities, 32, 108–119.
http://dx.doi.org/10.1177/002221949903200202
Spelke, E. (2003). What makes us smart? Core knowledge and natural language. In
D. Genter, & S. Goldin-Meadow (Eds.), Language in mind (pp. 277–311).
Cambridge, MA: MIT Press.
Spelke, E. S., & Tsivkin, S. (2001). Language and number: a bilingual training study.
Cognition, 78, 45–88. http://dx.doi.org/10.1016/S0010-0277(00)00108-6
Starkey, P., Klein, A., & Wakeley, A. (2004). Enhancing young children’s
mathematical knowledge through a pre-kindergarten mathematics
intervention. Early Childhood Research Quarterly, 19, 99–120. http://dx.doi.org/
10.1016/j.ecresq.2004.01.002
Starr, A., Libertus, M. E., & Brannon, E. M. (2013). Number sense in infancy predicts
mathematical abilities in children. Proceedings of the National Academy of
Sciences of the United States of America, 110, 18116–18120. http://dx.doi.org/10.
1073/pnas.1302751110
Toll, S. W. M., & Van Luit, J. E. H. (2014a). Explaining numeracy development in
weak performing kindergartners. Journal of Experimental Child Psychology, 124,
97–111. http://dx.doi.org/10.1016/j.jecp.2014.02.001
Toll, S. W. M., & Van Luit, J. E. H. (2014b). The developmental relationship between
language and low early numeracy skills throughout kindergarten. Exceptional
Children, 81, 64–78. http://dx.doi.org/10.1177/0014402914532233
Vandermaas-Peeler, M. (2008). Parental guidance of numeracy development in
early childhood. In O. Saracho, & B. Spodak (Eds.), Contemporary perspectives on
mathematics in early childhood education (pp. 277–290). Charlotte, NC:
Information Age.
Vandermaas-Peeler, M., Nelson, J., Bumpass, C., & Sassine, B. (2009).
Numeracy-related exchanges in joint storybook reading and play. International
Journal of Early Years Education, 17, 67–84. http://dx.doi.org/10.1080/
09669760802699910
Vgotsky, L. S. (1978). Mind and society: the development of higher psychological
processes. Cambridge. MA: Harvard University.
Wynn, K. (1990). Children’s understanding of counting. Cognition, 36, 155–193.
http://dx.doi.org/10.1016/0010-0277(90)90003-3
Wynn, K. (1992). Children’s acquisition of the number words and the counting
system. Cognitive Psychology, 20, 220–251. http://dx.doi.org/10.1016/0010-
0285(92)90008-P
Xu, F., & Spelke, E. S. (2000). Large number discrimination in 6-month-old infants.
Cognition, 74, B1–B11. http://dx.doi.org/10.1016/S0010-0277(99)00066-9