International Journal of Educational Research Open 2–2 (2021) 100035

Contents lists available at ScienceDirect

International Journal of Educational Research Open

journal homepage: www.elsevier.com/locate/ijedro

Systematic review on the precursors of initial mathematical performance

Camila Peres Nogues∗, Beatriz Vargas Dorneles
Faculty of Education, Universidade Federal do Rio Grande do Sul, Brazil

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

Keywords: This study carried out a systematic review to identify which cognitive skills of general and specific domain are
Mathematical achievement most frequently indicated as predictors of mathematical achievement of primary school students. For that, 62
Cognitive skills studies were included from the EMBASE, American Psychological Association, PubMED/MEDLINE, and Educa-
Predictors
tional Resources Information Center/ERIC databases. Results indicated that working memory and early numerical
skills are the cognitive abilities most often reported as predictors of later mathematical achievement, regardless
of the school grade and the type of mathematical performance considered. The findings gather evidence that
contributes to the understanding of which cognitive skills are fundamental for long-term mathematics learning
and that such skills can be inserted in school education as a resource to prevent later mathematical difficulties.

1. Introduction These findings are corroborated by other studies which focused

on the evaluation of first-grade students that also highlight working
Learning Mathematics requires a series of base cognitive abilities, memory and early numerical abilities as predictors of later mathe-
general and specific, that determine if this process will be easy and quick matical performance (e.g. Aragón et al., 2019; Nunes et al., 2007;
to students. Therefore, it is necessary to identify which cognitive abili- Passolunghi, Vercelloni, & Schadee, 2007; Xenidou-Dervou et al., 2018).
ties are involved in mathematics learning, to help teachers’ work and, However, a British study which evaluated 77 students from 7.8 to 10.8
consequently, students’ learning, thus minimizing future problems. The years old, has also considered abilities from general and specific do-
development of interventions and teaching strategies that can effectively mains (Gilmore et al., 2018). It shows, among the evaluated abilities,
improve students’ mathematical learning depend on the identification only number line estimation as a predictor of mathematical perfor-
of abilities of early quantitative knowledge that influence later develop- mance, having no meaningful results to any of the general domain abil-
ment (Geary, 2011). ities, working memory among them. This study meets other works that
In this sense, longitudinal studies have identified the relation of early have also investigated the abilities of both domains and indicated dif-
numerical abilities and general cognitive abilities with later mathemati- ferent results (e.g. Cirino, 2011; Fanari et al., 2018; Hawes et al., 2019;
cal achievement (e.g. Carr et al., 2020; Gashaj et al., 2019; Geary, 2011; Orrantia et al., 2018).
LeFevre et al., 2010; Xenidou-Dervou et al., 2018). An American longi- It is worth noting that, in studies such as the one by Jordan and
tudinal study (Geary, 2011) evaluated general and specific abilities and colleagues (2010), in which the performance of the same students was
their effects on the mathematical achievement of 177 children from the evaluated in more than one school grade, the authors found different
1st to the 5th grade. From the results, it can be identified that process- results on the predictor abilities in each grade. This study, also from the
ing speed, working memory, recognition of quantities, counting, and U.S., aimed to longitudinally evaluate the unique and predictive contri-
number estimation were abilities that contributed to the development bution of number sense and general abilities, such as working memory
of mathematical knowledge during these school grades. The findings of and spatial reasoning, in 175 students in 1st grade, who were later eval-
this study indicated that these abilities were also identified as having an uated in 3rd grade. From the results, they identified that number sense
explanatory value of mathematical performance, that is, when evaluated and spatial reasoning were predictors in both grades, however work-
early in the school life, they predict later mathematical achievement. ing memory was explicative of mathematical achievement only on 3rd
From these evidences, the study concluded that mathematical perfor- graders. Thereby, we can raise the hypothesis that the abilities that influ-
mance and its development through time are boosted by a combination ence mathematical achievement depend on the school grade evaluated,
of general domain abilities, which affect several learning areas, with as well as the type of mathematical performance considered, if arith-
abilities from a specific domain, which influence only mathematical metical or problem solving, for example (Kiss et al., 2019; LeFevre et al.,
learning (Geary, 2011). 2010; Toll et al., 2016).

∗
Corresponding author.
E-mail addresses: [email protected] (C.P. Nogues), [email protected] (B.V. Dorneles).

https://doi.org/10.1016/j.ijedro.2021.100035
Received 30 November 2020; Received in revised form 14 January 2021; Accepted 15 January 2021
Available online 29 January 2021
2666-3740/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
C.P. Nogues and B.V. Dorneles International Journal of Educational Research Open 2–2 (2021) 100035

Thus, considering the divergences and convergences among the find- the references of a manual research. The search comprised articles pub-
ings of the researches in the same themes, we highlight the importance lished until March 3rd , 2020, without establishing an initial limit, as we
of gathering and analyzing the existing studies to check the abilities intended to encompass the highest number of studies on the subject. The
more frequently related as precursors of mathematical performance. search strategies were the same in all bases, considering the key words:
Therefore, this defines the proposal of this study which aims to do a (predictors) OR (precursors) AND (mathematics) AND (achievement). The
systematic review of the literature on cognitive abilities identified as key word mathematics was considered because most authors used this
predictors of mathematical performance, especially arithmetical perfor- term to also designate arithmetic performance, the focus of this system-
mance, of elementary students. The intention of this systematic review atic review. According to previously established criteria, we selected
is the possibility of mapping the literature presented until now about only complete articles published in peer-reviewed journals, in English,
the precursors of mathematical performance, i.e., which abilities, from Spanish or Portuguese, that aimed to investigate cognitive abilities that
general and specific domains, that are considered as explanatory for this could predict mathematical performance in Elementary education.
performance. It is relevant to mention that this systematic review was
conducted considering the perspective of cognitive psychology, specif- 2.2. Criteria for eligibility
ically numerical cognition. Thus, all articles included for analysis are
on the same theoretical foundation and consider methods for assessing At the first phase, screening, the articles were selected through read-
student performance that are in accordance with the perspective men- ing the titles and abstracts, including the studies that were in agreement
tioned. with the topic to be studied and excluding those that were repeated
The focus on elementary students was determined from the under- or had no relevance to the theme. After, we did a list of eligible ar-
standing of how mathematical learning takes place, which requires that ticles, which were partially read, considering mainly the method, the
the basic abilities are well consolidated to have an adequate develop- sample, evaluated precursor abilities, and the outcome variables con-
ment in this area of school knowledge (Butterworth, 2005; Nunes & sidered. In this eligibility phase, we followed the ensuing criteria for
Bryant, 1997). Besides this, we highlight the relevance to look for re- inclusion: a) sample of students from Preschool and early grades of El-
searches that present regression analysis and that describe the results ementary school; b) sample with typical performance; c) evaluation of
from the point of view of predicting mathematical achievement. How- cognitive abilities of general or specific domains of learning; d) conduct
ever, we should note that mathematical performance is broad and in- an analysis of prediction (regression analysis); and e) mathematical per-
volves several areas, such as arithmetic, algebra, problem solving, and formance, related to arithmetic performance, as outcome variable. The
geometry, among other. Therefore, as the age range chosen for this sys- exclusion criteria were: a) sample smaller than 15 participants; b) eval-
tematic review comprises students from Preschool until the 5th grade, uation exclusively of social, psychological, or emotional measures; c)
we opted to select studies that included arithmetic as the main area to evaluation of other abilities (for example: mental activity, physical ac-
evaluate mathematical achievement, as it is the central axis of the math- tivities, fine motor skills, spontaneous focus on numbers); d) evaluation
ematical curriculum in this schooling phase (Brasil, 2018), dealing with of a teaching method or teacher training; e) evaluation based on parental
the study of numbers, operations, and their relations. Thus, we intend level of education or school incentive at home; f) sample composed by
to gather evidence of cognitive abilities that can help students’ early students with learning difficulties, disorders, or syndromes; and g) out-
mathematical development. comes different from those of arithmetic performance or which included
In this sense, knowing the different studies that investigate the pre- reading or problem solving.
cursors of mathematical performance is a key to identify what we al-
ready know on the theme and what types of knowledge can be broaden 2.3. Criteria for inclusion and data selection
by future researches that contribute to the identification of important
abilities for mathematical performance. This allows the evaluation of After that, in the inclusion phase, we read the complete studies se-
those abilities in children with and without mathematical difficulties lected to extract the data. The data of interest were organized in a
and plan interventions that help the development of those abilities and, spreadsheet, considering the following information: a) publication year;
consequently, of mathematical knowledge. b) country in which the study was held; c) the main objective of the
study; d) characteristics of the sample (age and schooling level); e) eval-
2. Method uation moments; f) evaluated cognitive abilities; g) outcome variable
considered; h) main results.
This study of systematic review of literature was conducted and re- To analyze this data, we considered: the frequency of countries in
ported according to the recommendations of Preferred Reporting Items which the studies originated, the average, the median, and the interval
for Systematic Reviews and Meta-Analyses – PRISMA (Moher et al., 2015), of the sample size; the school grades evaluated; the most and least fre-
which contains a set of necessary items to be followed by the authors, quent cognitive abilities evaluated; the outcome variables considered;
helping them to conduct a standard method to elaborate a systematic and, according to the main results, the most frequent predictors to arith-
review. PRISMA is widely used on researches in the health area, there- metic performance.
fore, some items do not apply to a review of education studies, even so,
due to the lack of methods devised for the educational area, we opted to 3. Results
adopt these existing recommendations, adapting when necessary. Thus,
we followed the phases of identification, screening, eligibility, inclusion, Initially, from the search in the database, we identified 1991 articles
and data extraction from the included studies. All phases were done by and included another 28 from a manual search, in a total of 2019 arti-
the researcher rigorously following the pre-established criteria. We did cles (Fig. 1). After the triage, we read the titles and abstracts, excluding
not register the protocol of this systematic review, as the researched the repeated ones and those with no relation to the proposed theme,
topic is not on the scope of the registry platform. such as those whose outcome variable was other than arithmetic perfor-
mance, those that did not evaluate cognitive abilities, and those which
2.1. Search strategy did not focus on the prediction analysis, or whose sample was for ad-
vanced school years, or with children with learning difficulties. This left
The data bases EMBASE, American Psychological Association (APA), us with 217 studies to be read in more details. In the eligibility phase,
PubMED/MEDLINE, and Educational Resources Information Center (ERIC) we excluded 155 articles, according to the exclusion criteria previously
were chosen as they were relevant to the subject area we intended to mentioned. Thus, we had 62 studies left to read in full and synthetize in
study and academically relevant. We also included articles selected from a qualitative analysis.

2
C.P. Nogues and B.V. Dorneles International Journal of Educational Research Open 2–2 (2021) 100035

Fig. 1. PRISMA flow diagram of systematic re-

view.
Source: Adapted from PRISMA (Moher et al.,
2009).

Graph 1. Frequency of articles per publication

year.
Source: Created by the author.

From this selection of 62 articles included in this systematic review, saying that the median of sample sizes was 148 students, 17 the smaller
it was possible to identify that the period of publication of these stud- sample and the second larger 781 students, not considering the biggest
ies was between 2001 and 2020, most of them in the year 2019 (n=7, of all which was the cohort study previously mentioned. The average
11.3%), as can be seen in Graph 1. Regarding the journals, the selected age of participants, considering the start of data collection in all studies,
articles were distributed through 31 journals, the ones with the highest was 6.11 years old. The different age averages presented in the studies
number of studies published were Journal of Experimental Child Psychol- encompassed the interval from 3.8 to 10.3 years old.
ogy (n=11, 17.7%), followed by Journal of Educational Psychology (n=7, Most studies focused on the evaluation of children in Preschool
11.3%), and Developmental Psychology (n=5, 8.1%). (n=38, 61.3%). Longitudinal studies were the most common in this re-
The data was collected in 15 countries, most studies were held in view (n=32, 51.6%), with the evaluation of mathematical performance
the United States (n=24, 38.7%), followed by the United Kingdom (n=8, done after a year of predictor abilities, and 30 studies (48.4%) in which
12.9%), Belgium (n=5, 8.1%), Spain, Netherlands, and Italy each with 4 the outcome was measured during the same school year as the other
studies, followed by Canada, China, Finland, and Sweden with two stud- cognitive abilities. Students’ school level varied from Preschool as the
ies each, and, finally, Germany, Australia, Cuba, France, and Switzer- starting point, in which the predictor abilities were evaluated, until the
land with one study per country (Graph 2). The average sample was 5th grade, to evaluate the outcome variable.
363.2 participants; however, this was due to some cohort studies, one Out of the analyzed studies, 41 (66.1%) used only standardized tests
specially had a sample size of 12,099 students. Therefore, it is worth as measure of mathematical performance, 12 (19.35%) used only in-

3
C.P. Nogues and B.V. Dorneles International Journal of Educational Research Open 2–2 (2021) 100035

Graph 2. Distribution of publications per

country.
Source: Created by the author.

formal tasks of research, and 8 (12.9%) measured mathematical perfor- sidered it as part of the number sense that also did not separated it on
mance from a combination of standardized tests and informal tasks. We their analysis (e.g. Jordan et al., 2007).
also saw that only 15 articles (24.2%) analyzed only arithmetic perfor- From the analysis, we can say that in the first grades of schooling,
mance, the other 47 (75.8%) measured students’ general mathematical from Preschool until the 2nd grade, more basic abilities are essential,
performance, including number knowledge, arithmetic, problem solving as approximate number system (ANS), recognition and reading of num-
and, in some even notion of geometry, for example. Table 1 presents a bers, and identification of quantities (e.g. Cirino, 2011; Gimbert et al.,
descriptive synthesis of the selected studies, in which more details can 2019; Mazzocco et al., 2011; vanMarle, Chu, Li, & Geary, 2014). As
be seen. students advance the school grades, mathematical learning will depend
We also found that most (n=48, 77.4%) articles evaluated measures of more complex abilities, such as arithmetic knowledge, which in-
of specific and general domain, 5 (8.1%) evaluated only abilities of the clude numerical operations (addition and subtraction), arithmetic facts,
specific domain, and 9 (14.5%) only the general domain. From this syn- calculation from algorithms or mental strategies (Casey et al., 2017;
thesis of studies, we could see that the specific domain abilities that Gilmore et al., 2018).
more frequently predict mathematical performance are: early numeri- However, some abilities are important predictors during the whole
cal abilities – serialization, identification and naming numbers, numeric schooling period and are shared among all levels of mathematical
sequence, discrimination of quantities; counting – cardinality, counting knowledge, which are: counting, number estimation, and early numer-
strategies; and comparison of symbolic quantities. Similarly, the abili- ical abilities. This last one includes number recognition (understanding
ties of general domain are: working memory, intelligence (verbal and of the number system), comparison of symbolic amounts, number sense,
non-verbal); executive functions, and phonological awareness. which was identified as a predictor of mathematical performance up the
Another interesting factor to be highlighted are the differences and 3rd grade, and subitizing, seen in the performance of 4th and 5th grade
similarities found in the predictor variables of arithmetic performance. students. Some differences between explanation values and the level of
Out of all articles included, 28 (45.2%) evaluate working memory and significance of the evaluated abilities also appeared according to the
out of those, 21 (75%) found that it was a predictor. The second ability tasks used, as well as the theoretical definitions followed by the authors
most frequently evaluated was students’ intelligence, 28 articles (45.2%) on each cognitive ability.
considered this ability and 14 (50%) identified it as a predictor of math- Regarding general domain abilities, we can see that most of the ones
ematical achievement. When checking the specific domain abilities, out chosen to be evaluated are demanded through the first stages of school-
of the 26 studies (41.9%) which considered early numerical abilities in ing, that is, since Preschool until the 5th grade, varying only on the
their evaluation, 25 (96.15%) reported that these measures explained choice of tasks used or the study focus which preferred to evaluate a
later mathematical achievement. Only one study (Fuhs et al., 2016) certain ability of a set of abilities. Hence, it was not possible to iden-
which evaluated early abilities did not find a significant predicting tify any tendency of predicting abilities that are more recurrent in each
value. This probably happened because in the regression model analy- school grade, as the abilities were important and explanatory of stu-
sis these variables were inserted together with the general mathematical dents’ mathematical performance through all schooling years. The abili-
performance, which was also measured in the first evaluation moment ties were: working memory, phonological awareness (except 2nd grade),
and this evaluated performance in Preschool was responsible for ex- intelligence quotient (IQ), some type of spatial ability, be it on memory,
plaining a large part of the later mathematical achievement, evaluated orientation or attention, and processing speed.
in the 2nd grade (effect size 𝛽=0,48). In the sequence of analysis of this Similarly, these abilities also vary their explanatory level depending
study, the authors show the indirect effects of early numerical abilities, of what is evaluated as mathematical performance, i.e., when there
which were responsible for mediating the direct relation between previ- is a distinction between only arithmetic performance, which includes
ous mathematical performance (Preschool) and later performance (2nd calculations, algorithms, fluency, and the general mathematical per-
grade). formance, which considers, besides the ability to solve calculation,
About counting, the second most frequent specific ability in the stud- problem solving, measurements, geometry, algebra, among others.
ies, it was a predictor in 12 (70.6%) studies out of the 17 (27.4%) which Working memory and early numerical abilities were the cognitive abili-
included it in their analysis. Here it is worth to mention that some stud- ties more frequently related as predictors in both types of performance;
ies considered counting as part of the early abilities, not separating it followed by symbolic and non-symbolic comparison, and language
on their analysis (e.g. Jordan et al., 2009), as well as studies which con- abilities (writing, reading words, verbal fluency, vocabulary) for the

4
 C.P. Nogues and B.V. Dorneles
Table 1
Summarized description of each study selected for the systematic review.

Study (IF) N Age (average) Evaluated school grade Specific domain abilities General domain abilities Outcome variables Significant predictors and effect size (𝛽)

Hecht et al., 201 7.7 years old From the 2nd to the Reading, phonological Calculation abilities: (four operations, fraction, 2nd to 3rd grade and 2nd to 5th
2001 (IF = 2.84) 5th grade memory, phonological algebra - WJ-III); simple arithmetic (mental grade: phonological memory (0.05
awareness, naming speed, calculation) and 0.03), rate of access (0.12 and
IQ 0.07), and phonological awareness
(0.06 and 0.11).
3rd to 4th grade: phonological
awareness (0.06).
4th to 5th grade: phonological
awareness (0.02).
Aunola et al., 194 6.25 years old Preschool, 1st, and Counting ability Visual attention, Mathematical performance: Diagnostic Test for Counting ability (0,62); metacognitive
2004 2nd grades metacognitive knowledge, Basic Mathematical Concepts – Standardized test. knowledge (0,17); listening
(IF = 4.88) listening comprehension Considered: knowledge of ordinal and cardinal comprehension (0.16).
numbers, basic mathematical concepts,
identification of numbers, simple problems, basic
arithmetic
Fuchs et al., 312 Not informed 3rd grade Non-verbal problem Language; concept Arithmetic (fluency in calculations of addition Arithmetic: attention (0.24);
2006 solving formation, processing and subtraction); calculations with two-digit processing speed (0.33); and
(IF = 3.82) speed; long-term memory; numbers; arithmetic problems with numbers up phonological decoding (0.16)
working memory; to 9 Calculations with two-digit
attentive behavior; numbers: attention (0.37).
phonological decoding and Arithmetic problems: non-verbal
reading problems (0.16), concept formation
(0.20), reading (0.11), language (0.18)
5

and attention (0.19).

Nunes et al., 53 6 years old 1st grade (later, 2nd Logic competence BAS II – general cognitive SATs-Maths (standardized test): numerical Logic competence (0.45); counting
2007 grade) (knowing how to read and abilities (quantitative knowledge, basic operations recall (0.12); intelligence - BAS II
(IF = 1.03) write numbers, ability; verbal similarities; (0.39).
understand number numerical and matrices
relations and operations) abilities); working
memory (counting recall
and backward digit recall)

International Journal of Educational Research Open 2–2 (2021) 100035

Passolunghi et 170 6.3 years old 1st grade Numerical competence Working memory; Standardized test for the 1st year (logic, Working memory (0.38), specifically
al., 2007 (writing and reading short-term memory; arithmetic, and geometry) central executive, and verbal counting
(IF = 2.65) numbers, comparing phonological ability; IQ (-0.54).
magnitudes, counting
ability)
Jordan et al., 277 5.58 years old Preschool and 1st Number sense (counting, Reading (letter naming Mathematical achievement: WJ – III (standardized Number sense (0.72).
2007 grade numerical knowledge, fluency, segmentation test) – calculation and mathematical problems
(IF = 4.51) non-verbal calculation, fluency and nonsense
arithmetic problems, word fluency)
number combination)
Andersson, 141 10.3 years old 3rd and 4th grades Reading and Arithmetic (addition, subtraction, and Arithmetic: counting span (0.24);
2008 comprehension. multiplication) and arithmetic fact retrieval verbal fluency (0.14); trial-making
(IF = 1.42) Working memory (central (-0.19); digit span (0.17); age (0.15);
executive, phonological IQ (0.18) and reading (0.30).
loop, and visuospatial Recovery of arithmetic facts: reading
sketchpad) (0.34); verbal fluency (0.19) and
trail-making (-0.17).
(continued on next page)
 C.P. Nogues and B.V. Dorneles
Table 1 (continued)

Study (IF) N Age (average) Evaluated school grade Specific domain abilities General domain abilities Outcome variables Significant predictors and effect size (𝛽)

Bull et al., 104 4.5 years old Preschool (later, 1st Central executive tasks, Performance Indicators in Primary School (PIPS) – In the beginning of 1st grade:
2008 and 3rd grades) short-term working evaluates basic number knowledge, phonetic, and Short-term visuospatial memory
(IF = 12.22) memory and working reading ability (r2 =8.3%), short-term memory
memory (r2 =2.9%), verbal work memory
(r2 =7.1%), inhibitory control (r2 =3.3%)
and planning and monitoring (r2 =5%).
In the end of 1st grade: Short-term
visuospatial memory (r2 =2,7%)
In the endo f 3rd grade: visuospatial
working memory (r2 =5.5%)
Locuniak; 198 5.5 years old Preschool and 2nd Number sense (counting, Reading (letter-naming Fluency in calculation (Assessment of Math Fact R2 =42%.
Jordan, 2008 grade number knowledge, fluency, Fluency) – addition up to 18, and subtractions Number sense – number knowledge
(IF = 2.53) nonverbal calculation, phoneme-segmentation with minuends up to 18) (0.17), number combinations (0.28)
story problems, and fluency, and Working memory – backward digit
number combinations) nonsense-word fluency) span (0.14).
Cognitive measurements:
digit span (WISC-IV) and
IQ (WASI – vocabulary and
nonverbal reasoning)
Simmons et al., 42 5.2 years old Not informed Visuospatial functioning, Arithmetic: British Ability Scales Number Skills R2 =41% - visuospatial functioning
2008 (outcome evaluation phonological awareness, Test (BAS Number) standardized test that (0.32) and phonological awareness
(IF = 0.89) done after 1 year) vocabulary, nonverbal evaluates the solution of simple arithmetic (0.30).
reasoning, reading problems
De Smedt et 106 6.25 years old 1st and 2nd grades Working memory (central Mathematical performance (Flemish Student 1st grade: R2 =42% - visuospatial
6

al., 2009 executive, phonological Monitoring System – standardized test) – number sketchpad (0.34), central executive
(IF = 3.26) loop, and visuospatial knowledge, comprehension of operations, simple (0.26), nonverbal IQ (0.21).
sketchpad) and nonverbal arithmetic, and measurements 2nd grade: R2 =49% - phonological
IQ loop (0.24), nonverbal IQ (0.42)
De Smedt et 42 6.3 years old 1st grade (later, 2nd Numerical comparison, Nonverbal IQ Mathematical performance: Flemish Student Number comparison (0,34).
al., 2009 grade) reading numbers Monitoring System (standardized test). Evaluates
(IF = 3.20) the mastery of numerical system up to 20,
numerical knowledge, understanding of

International Journal of Educational Research Open 2–2 (2021) 100035

operations, simple arithmetic, problems, and
measurements
Desoete et al., 158 Not informed 1st , 2nd and 3rd TEDI-MATH: seriation, Arithmetic performance: Kortrijk Arithmetic Arithmetic performance – 3rd grade:
2009 grades (later, 3rd , 4th , classification, procedural Test–Revision - KRT-R (standardized test). conceptual counting (0.45) and
(IF = 6.44) 5th grades) counting knowledge, Evaluates mental arithmetic and numerical numerical ability (0.55), 4th grade:
conceptual counting knowledge. conceptual counting (-0.22) and
knowledge Numerical ability: Arithmetic Tempo Test (ATT), numerical ability (0.22), 5th grade:
in which the child has to solve the highest numerical ability (0.34).
number of numerical facts (4 operations) possible Numerical ability – 3rd grade:
in 5 minutes. procedural counting (0.54) 5th grade:
procedural counting (0.30)
Jordan et al., 196 5.5 years old Preschool (later, 1st Number competence: Mathematical performance: WJ-III (standardized Number competence (0.96)
2009 and 3rd grades) counting, number test). Sub-test calculations (WJCalc) measures the
(IF = 6.97) recognition, number ability to solve conventional calculations. Sub-test
comparison, nonverbal of applied problems (WJApp) which demands
calculation, story solving problems presented orally
problems, and number
combinations
(continued on next page)
 C.P. Nogues and B.V. Dorneles
Table 1 (continued)

Study (IF) N Age (average) Evaluated school grade Specific domain abilities General domain abilities Outcome variables Significant predictors and effect size (𝛽)

Krajewski; 130 6.25 years old Preschool, 1st, and Basic numerical skills: Nonverbal intelligence, Mathematical achievement: arithmetic fact Arithmetic fact retrieval at 1st grade:
Schneider, 4th grades number word sequence, phonological memory, retrieval and two tests (1st and 4th grades) QNC (0.35) and number naming speed
2009 counting; and QNC: speed of access to standardized in Germany (number line, addition, (0.30).
(IF = 3.07) quantity comparison and long-term memory subtraction, multiplication, division, practical Mathematical performance at 1st
quantity to number-word (number naming speed), mathematics, and geometry) grade: arithmetic fact retrieval (0.33),
linked seriation socioeconomic status basic numerical skills (0.42).
Mathematical performance at 4th
grade: number naming speed (0.26),
socioeconomic status (0.36), basic
numerical skills (0.49).
Aunio; 212 6 years old Preschool (later, 1st Early numeracy: relational Basic and applied arithmetic abilities: Basic numerical abilities: relational
Niemivirta, grade) skills – concepts of Mathematics school test (standardized test) – skills (0.30) and counting skills (0.24).
2010 comparison, classification, addition and subtraction up to two digits Applied arithmetic abilities:
(IF = 3.19) one-to-one Mathematical competence: evaluation done by relational skills (0.25) and counting
correspondence, seriation; teachers according to the school curriculum and skills (0.26).
counting skills – the use content studied until the moment Mathematical competence: relational
of number words, skills (0.24) and counting skills (0.47).
counting, and the general
understanding of numbers
Gilmore et al., 83 5.8 years old Preschool Non-symbolic approximate Verbal intelligence and Mathematical achievement: counting, symbolic Verbal intelligence (R2 =22.7%),
2010 addition, verbal knowledge test literacy level knowledge of numbers and some abilities of literacy (R2 =32.3%) and symbolic
(IF = 2.81) of numbers and geometry knowledge of numbers (R2 =11.1%).
knowledge of Arabic
7

numerals.
Jordan et al., 175 5.5 years old 1st grade (later, 3rd Number sense: Number Intellectual level (WASI) – Mathematical achievement: Woodcock-Johnson III 1st grade - IQ: vocabulary (0.14),
2010 grade) sense brief screen – vocabulary and spatial (WJ-III) – written calculation and applied spatial reasoning (0.17); working
(IF = 4.29) counting knowledge and reasoning. problem solving memory (0.11); number sense (0.48).
principles, number Short-term memory and 3rd grade - IQ: spatial reasoning
recognition, number working memory (digit (0.25); number sense (0.46).
comparisons, nonverbal span)
calculation, story

International Journal of Educational Research Open 2–2 (2021) 100035

problems, number
combinations
LeFevre et al., 122 Not informed Preschool (later, 2nd Quantitative knowledge: Linguistic skills: Mathematical knowledge: KeyMath Test– Revised Standardized test – Enumeration:
2010 grade) evaluate and discriminate phonological awareness, (standardized test) – enumeration, geometry, and spatial attention (0.14), linguistic
(IF = 4.96) quantities, recognize vocabulary, verbal measurement. And the calculation test (WJ–R). skills (0.54), quantitative skills (0.19).
quantities (subitizing) reasoning, listening Also applied the number line task, symbolic Calculation: spatial attention (0.27),
Early numeracy skills: comprehension magnitude comparison task and word reading linguistic skills (0.39), quantitative
number naming and Spatial attention: skills (0.20). Geometry: spatial
nonlinguistic arithmetic visual-spatial working attention (0.23), linguistic skills (0.41).
memory and central Measurements: spatial attention
executive (0.24), quantitative skills (0.32).
Number line estimation (precision):
gender (0.21), spatial attention (0.24),
linguistic skills (0.47), quantitative
skills (0.24).
Comparison of magnitudes: linguistic
skills (0.24) and quantitative skills
(0.30).
 C.P. Nogues and B.V. Dorneles
Table 1
(continued)

Study (IF) N Age(average) Evaluated school Specific domain abilities General domain abilities Outcome variables Meaningful predictors and effect size (𝛽)
grades

Cirino, 2011 286 6.13 years old Preschool Symbolic and nonsymbolic Visuospatial working Addition (small sums) Symbolic comparison (0.04), symbolic
(IF = 1.42) comparison, symbolic memory, phonological labeling numbers (0.49), rote counting
labeling numbers, rote awareness, rapid (0.04) and counting knowledge
counting (numerical automatized naming (RAN) (-0.30).
sequence), counting
knowledge (count objects)
Geary, 2011 177 6.16 years old Preschool (later, 1st Counting knowledge; Intelligence (IQ): WASI Numerical operations of WASI (Wechsler Intelligence (0.24), processing speed
(IF = 6.68) grade) addition strategy choices; (vocabulary and matrix Individual Achievement Test–II– Abbreviated) (0.26), central executive (-0.29),
number sets (establish sets reasoning); working visuospatial sketchpad (0.26), number
with a determined memory (central sets, addition strategies, number line
quantity); number line executive, phonological estimation (0.14).
estimation loop, visuospatial
sketchpad), processing
speed
Libertus et al., 174 4.2 years old Preschool Approximate Number Verbal ability – vocabulary Mathematical ability: Test of Early Mathematics ANS (R2 =19%)
2011 System (ANS) — Ability (TEMA-3) – Evaluates numerical abilities
(IF = 4.67) non-symbolic numerical (counting, number comparison, reading numbers,
comparison comprehension of numerical facts, calculations of
addition and subtraction, numner concepts)
Mazzocco et 17 From 3 to 6 Preschool (later, ANS Intelligence - Wechsler Mathematical performance: Test of Early ANS (R2 =28%)
8

al., 2011 years old some still in Abbreviated Scale of Mathematics Ability (TEMA-3)
(IF = 3.97) Preschool, other in Intelligence (WASI) and
1st and 2nd grades) Rapid Automatized
Naming (RAN)
Passolunnghi; 70 5.2 years old Preschool (later, 1st Verbal counting, number IQ, working memory, Mathematical achievement: standardized test for Number competence evaluated in
Lanfranchi, grade) competence (comparison, verbal and visuospatial 1st grade that evaluates arithmetic and geometry the end of preschool: processing
2012 classification, short-time memory, speed (0.46) and working memory
(IF = 2.25) correspondence phonological ability and (0.38).

International Journal of Educational Research Open 2–2 (2021) 100035

one-to-one, seriation, processing speed Mathematical achievement
counting, use of evaluated at the end of the 1st
words-numbers, general grade: processing speed (0.23),
understanding of number) number competence (0.24) and verbal
IQ (0.29).
Sasanguie et 72 Preschool: 5.6 Preschool,1st and 2nd Number identification Writing: Flemish Student Mathematical performance: Flemish Student School grade (0,44), writing (0.35),
al., 2012 years old grades (number priming), number Monitoring System Monitoring System (standardized test). Evaluates symbolic number comparison (0.61),
(IF = 1.10) 1st : 6.7 years comparison, and symbolic (standardized tests), number knowledge, understanding of operations, non-symbolical number line
old and non-symbolic number evaluates the spelling of simple arithmetic, problem solving, estimation (0.38).
2nd : 7.6 years line estimation letters, words, and phrases measurements, and geometry
old
Nyroos; 40 From 8 to 9 3rd grade Working memory: central Mathematical performance: national standardized Visuospatial sketchpad (0.35) and
Wiklund- years old executive, phonological test (arithmetic calculation, mental arithmetic, phonological loop (0.43).
Hornqvist, loop and visuospatial time, understanding of numbers, fractions, area,
2012 sketchpad and volume)
(IF = 0.31)
(continued on next page)
 C.P. Nogues and B.V. Dorneles
Table 1 (continued)

Study (IF) N Age(average) Evaluated school Specific domain abilities General domain abilities Outcome variables Meaningful predictors and effect size (𝛽)
grades

Van der Ven et 211 6.5 years old 1st grade (later, 2nd Executive functions: Mathematical performance: standardized test Updating (0.84)
al., 2012 grade) inhibition, shifting and applied by the schools (numbers and relations,
(IF = 3.04) updating simple addition and subtraction, simple
multiplication and division, math applications,
and measuring)
Libertus et al., 204 4.8 years old Preschool Approximate Number Expressive vocabulary; Mathematical ability: Test of Early Mathematics ANS (R2 =7%), number abilities
2013 System (ANS) — attention; response time; Ability (TEMA-3). Evaluates number abilities evaluated in the beginning of the
(IF = 2.65) non-symbolic number memory span (digits and (counting, number comparison, reading numbers, school year (R2 =45%), response time
comparison letters) understanding of number facts, calculations of (R2 =6%).
addition and subtraction, number concepts)
Reigosa-Crespo 49 9.3 years old 3rd and 4th grades Basic numerical abilities: Non-verbal IQ, word and Mathematical fluency, curricular mathematical Mathematical fluency: age (0.36),
et al., 2013 (later, 4th and 5th dot enumeration, symbolic pseudoword reading, performance non-verbal reasoning (0,32), subitizing
(IF = 0.39) grades) comparison, simple reading fluency and (0.29).
reaction time; counting reading comprehension Mathematical performance based on
effect and subitizing effect curriculum: non-verbal reasoning
(0.68), subitizing (-0.33).
9

Sasanguie et 71 1st: 6.6 years 1st , 2nd and 3rd Symbolic and Writing: Flemish Student Arithmetic fluency and general performance in Arithmetical fluency: symbolic
al., 2013 old. 2nd : 7.7 grades (later, 2nd , 3rd non-symbolic comparison, Monitoring System Mathematics – Flemish Student Monitoring comparison (-0.36) and writing (0.38).
th
(IF = 3.25) years old and 4 grades) symbolic and (standardized test), System (standardized test). Evaluates numerical General mathematical performance:
3rd : 8.6 years non-symbolic number line evaluate spelling letters, knowledge, understanding of operations, simple symbolic comparison (-0.30), symbolic
old estimation words, and phrases arithmetic, problem solving, measures, and number line estimation (-0.27) and
geometry writing (0.33).
Bartelet et al., 209 6 years old Preschool (later, 1st Symbolic and Baseline response time; Arithmetic performance: standardized test (TTA). Symbolic comparison (0.28),
2014 grade) non-symbolic comparison, non-verbal reasoning Evaluates the level of automatization of basic estimation (0,30), gender (-0.12),

International Journal of Educational Research Open 2–2 (2021) 100035

(IF = 1.36) dot enumeration, artithmetic facts (they used: addition, school (-0.35).
estimation subtraction)
Fuchs et al., 394 6.5 years old 1st grade (later, 2nd Foundational numerical Non-verbal reasoning; Numeration understanding: KeyMath Non-verbal reasoning (0.09), listening
2014 and 3rd grades) competencies: number listening comprehension; –Numeration (identify amount of objects, reading comprehension (0.09), working
(IF = 0.64) sets (number, numerical working memory, and numbers, identify ordinal numbers, successor and memory (0.13), attentive behavior
mapping, and part-whole attentive behavior antecessor, order sequence of numbers, identify (008) and number sets (0.15).
relation) and number line numbers from the representation of base 10,
estimation identify numbers missing in the number line).
Calculation ability: Wide Range Achievement Test
– Arithmetic. Evaluates the ability to solve
arithmetic calculations
(continued on next page)
 C.P. Nogues and B.V. Dorneles
Table 1 (continued)

Study (IF) N Age(average) Evaluated school Specific domain abilities General domain abilities Outcome variables Meaningful predictors and effect size (𝛽)
grades

Martin et al., 193 6.16 years old Preschool (later, 1st Counting (procedural and Spatial working memory, Fluency in calculation (simple addition and Fluency: Verbal working memory
2014 grade) conceptual), symbolic phonological/verbal subtraction), computations (WJ-III - addition and (0,04), identification of numbers
(IF = 0.89) numbers (identification working memory, central subtraction with single and multi digits) applied (0.11), symbolic comparison (0.02).
and comparison) executive, phonological problems (WJ-III) and math problem solving Calculation: Age (-0.15), phonological
awareness, behavioral awareness (0.04), identification of
inattention numbers (0.1) and symbolic
comparison (0.02).
Applied problems: verbal working
memory (0.57), phonological
awareness (1.28), symbolic
comparison (0.23).
Problem solving: spatial working
memory (0.26), phonological
awareness (0.33), identification of
numbers (0.24), symbolic comparison
(0.09).
Szucs et al., 98 8.9 years old 3rd and 4th grades Number sense (symbolic Short-term memory: Mathematical performance: Mathematics Verbal intelligence (0.21),
2014 and non-symbolic verbal and visual; working Assessment for Learning and Teaching (MaLT – phonological decoding (0.27), visual
(IF = 1.58) comparison, subitizing); memory; phonological standardized test) and mathematical abilities – short-term memory (0.17), working
dot enumeration (from 4 decoding; verbal and Wechsler Individual Achievement Test (WIAT-II) memory (0.25), spatial orientation
to 6) non-verbal IQ; executive numerical operations (0.24) and executive functioning, and
10

functioning; switching; switching (trail-making) (-0.16).

attention; inhibitory
control; response time;
spatial orientation ability;
spatial knowledge of
symmetry; mental
rotation; finger
recognition; reading

International Journal of Educational Research Open 2–2 (2021) 100035

van Marle et 138 3.8 years old Preschool Quantity discrimination IQ, executive control, Mathematical performance: Test of Early ANS (0.28), counting (es = 0.20),
al., 2014 (ANS), counting, numeral preliteracy skills Mathematical Ability-3 (TEMA-3) numeral recognition (es = 0.81),
(IF = 1.99) recognition (1-15), (recognition of letters), numeral comparison (es = 0.04) and
numeral comparison, parental education cardinality (es = 1.93).
cardinality
Chu et al., 191 3.8 years old Preschool ANS and cardinality (IQ), executive functions, Mathematical performance: TEMA-3 – represent Phonological awareness (0.32), IQ
2015 (IF = 1.57) preliteracy (reading quantities with fingers, counting, number (0.19), cardinality (0.34).
ability) comparisons, informal arithmetic
Foster et al., 208 5.6 years old Preschool Fluid intelligence Numerical knowledge (numeracy): Numerical knowledge: IQ (0.88),
2015 (non-verbal IQ), Research-based Early Math Assessment (REMA) – phonological awareness (1.09)
(IF = 0.30) vocabulary, phonological verbal counting, counting objects, recognition of Applied problems: IQ (0.63),
awareness, RAN (Rapid numbers, and subitizing, number comparison, phonological awareness (0.68),
automatized naming), number sequence, recognition of numbers, short-term memory (0.59).
short-term phonological composition and decomposition of numbers,
memory addition, and subtraction.
Applied problems (WJ-III): simple arithmetic
problems
(continued on next page)
 C.P. Nogues and B.V. Dorneles
Table 1 (continued)

Study (IF) N Age(average) Evaluated school Specific domain abilities General domain abilities Outcome variables Meaningful predictors and effect size (𝛽)
grades

Geary, Van 197 3.8 years old Preschool Symbolic competences: Intelligence (IQ), executive Mathematical performance: Test of Early Non-verbal IQ (0.15), number
Marle, 2016 enumeration, verbal functions, preliteracy Mathematical Ability-3 (TEMA-3) – produce finger recognition (0.30), verbal counting
(IF = 0.98) counting, cardinality, (recognition of alphabet), patterns to represent different quantities, count, (0.20), ordinal comparison (0.11).
numeral recognition, parental education make numerical comparisons, simple arithmetic,
differentiate quantities. and number line
Non-symbolic
competences: ordinal
comparison, discrete
quantity discrimination
(ANS), continuous quantity
discrimination (AMS),
tracking small quantities,
nonverbal calculation
(implicit knowledge of
addition and subtraction)
Fuhs et al., 141 6.2 years old Preschool (later, 2nd Early number skills: Executive functions: Mathematical performance: Woodcock–Johnson III IQ (0.24), executive functions (0.25)
2016 (IF = 0.50) grade) quantity discrimination, inhibitory control and (WJ-III) – and mathematical performance
number line estimation, cognitive flexibility; IQ Applied problems (requires quantitative (WJ-III) also measured in preschool
number sets identification, reasoning, problem solving, and mathematical (0.48).
fast counting, number knowledge)
word comprehension
Nguyen et al., 781 4.3 years old Preschool (later, 5th Preschool mathematical Mathematical performance in the 5th year: Tools Counting and cardinality (0.42),
2016 grade) competences — for Elementary Assessment in Math 3–5 (TEAM geometry (0.13) and patterns (0.10).
(IF = 3.29) Research-based Early 3–5). Evaluates mathematical knowledge,
11

Mathematics Assessment including fractions, geometry, multiplication,

(REMA) – counting, division, and data interpretation.
number recognition,
addition and subtraction,
patterns, measurements,
geometric and spatial
competences
Purpura, Logan, 114 4.2 years old Preschool Mathematical language: Vocabulary: Test of Early numeracy: Preschool Early Numeracy Skills Mathematical language (0.48).

International Journal of Educational Research Open 2–2 (2021) 100035

2015 (IF = 0.97) Mathematical language Preschool Early Literacy Test – Brief Version (PENS-B) – set comparison,
subtest (assessing Skills (TOPEL) – numeral comparison, one-to-one correspondence,
comparative language – single-word spoken number order, number identification, ordinality,
more, less, take away; and vocabulary and the ability number combinations
spatial language – far, to formulate definitions
near, before); ANS for words; Executive
functions: inhibitory
control, attention, shifting,
and verbal working
memory
Tobia et al., Study From 3 to 6 Preschool (study 1 Number sense: (SNUP) – Battery of tests: Learning Calculation: addition and subtraction presented Study 1 (transversal): R2 =63% –
2016 (IF = 0.42) 1: 102 years old transversal, study 2 quantity comparison, Difficulties Indexes (IDA). verbally and non-verbally Vocabulary (0.11), size seriation
Study longitudinal with a counting, number line, Evaluates linguistic ability (0.13), visual-spatial memory (0.1).
2: 43 1-year interval) size seriation, semantic in preschoolers Study 2 (longitudinal): R2 =61.5% -
knowledge of digits, (articulation, vocabular, 0.09), size seriation (0.12).
visual-spatial memory phonological awareness,
morphosyntactic
comprehension, sequential
history, and knowledge of
letters
(continued on next page)
 C.P. Nogues and B.V. Dorneles
Table 1 (continued)

Study (IF) N Age(average) Evaluated school Specific domain abilities General domain abilities Outcome variables Meaningful predictors and effect size (𝛽)
grades

Casey et al., 138 6.7 years old 1st grade (later, 5th Verbal ability, arithmetic Spatial abilities Analytical mathematical reasoning, arithmetic Analytical mathematical reasoning:
2017 grade) (precision and strategies) fluency spatial abilities for 1st grade (0.52),
(IF = 0.69) arithmetic strategies – decomposition
(0.27), arithmetic precision (0.23),
arithmetic fluency 5th grade (0.34).
Arithmetic fluency: decomposition
strategy (0.46).
Ching, Nunes, 115 6.36 years old 1st grade (later, 2nd Counting and addition Working memory (central Mathematical performance: calculation and Calculation (R2 =28.8%; 58.9%; 58.9%),
2017 (IF = 0.42) grade) reasoning executive, phonological problem solving – based on the Chinese respectively): counting (R2
loop, visuospatial curriculum unique=4.1%), additive reasoning (R2
sketchpad, non-verbal IQ) unique=30%) and working memory
(R2 unique=11%).
Mathematical problems: (R2 = 58.1%):
additive reasoning (R2 unique=38.6%),
working memory (R2 unique=6.6%)
Gilligan et al., 12099 From 5 to 7 Not informed Cognitive and spatial Mathematical performance: National Foundation R2 = 42.5%.
2017 (IF = 1.39) years old ability: BAS II for Educational Research Progress in Maths (NFER Building of patterns at 5 years old
(standardized test) – PiM, standardized test – evaluates knowledge on (0.13), Building of patterns at 7 years
construction of patterns numbers, shapes, measurement, and data old (0.25), reading words (0.35),
(copy a pattern using a set interpretation vocabulary (0.12).
of building blocks, by
rotation and spatial
12

organization of blocks),
vocabulary, reading words,
and mathematics
Orrantia et al., 104 4.7 years old Preschool Comparison of Intelligence (non-verbal Mathematical performance: Test of Basic R2 =65.5%
2017 non-symbolic numerical IQ), processing speed, Mathematical Competence (TEMA-3) – evaluates Spatial memory (0.21), intelligence
(IF = 0.06) magnitudes, comparison of array of verbal memory, formal and informal concepts and abilities of (0.22), symbolic comparison (0.24),
symbolic magnitudes spatial memory, inhibitory arithmetic dot enumeration (0.23) and interval of
(1-9), dot enumeration control subitizing (0.24).

International Journal of Educational Research Open 2–2 (2021) 100035

Sánchez et al., 96 5.7 years old Preschool (later, 1st Comparison of small Visual-spatial memory, Additions (fluency), calculations (addition and R2 =59.6%
2017 (IF = 0.00) grade) symbolic magnitudes phonological memory, subtraction), numerical-verbal problems Spatial memory (0.25), symbolic
(1-9), rapid automatized non-verbal IQ, processing comparison (0.19), RAN (-0.26) and
naming (RAN) speed, phonological phonological awareness (0.26).
awareness (segmentation
of phonemes)
Wong, Ho, 153 6.1 years old Preschool (later, 1st Symbolic and Working memory Arithmetic problems, computations and Arithmetical problems: non-verbal
2017 (IF = 0.33) and 2nd grade) non-symbolic numerical (phonological loop, mathematical performance (Learning and reasoning (0.27) and phonological
processing; visuospatial sketchpad, Achievement Measurement Kit 2.0 Second Grade loop (0.21).
symbolic-nonsymbolic central executive); Mathematics (LAMK 2.0 Mb) – evaluates Calculation: magnitudes processing
mapping – magnitudes non-verbal reasoning; curriculum topic, arithmetic, shape and space, (symbolic and non-symbolic – 0.21),
processing reading words measurements, and simple statistic non-verbal reasoning (0.17), reading
words (0.23).
Mathematical performance:
computations (0.24) arithmetical
problems (0.30), non-verbal reasoning
(0.22).
(continued on next page)
 C.P. Nogues and B.V. Dorneles
Table 1 (continued)

Study (IF) N Age(average) Evaluated school Specific domain abilities General domain abilities Outcome variables Meaningful predictors and effect size (𝛽)
grades

Fanari et al., 43 6.5 years old 1st grade Early numerical Visuospatial working Mathematical achievement: AC-MT 6-11 – Test Early numerical competencies (0.50).
2018 (IF = 0.00) competence: lexical memory for the evaluation of calculating and
(knowledge of numerical problem-solving abilities. Evaluates operations
symbols), semantic (addition and subtraction), understanding of
(symbolic and quantities, of the synthetic structure of numbers
non-symbolic comparison), (tens and units), and semantic representation
counting, and (organize in order from the smaller to the bigger)
pre-syntactic
(correspondence
number-quantity)
Gilmore et al., 77 9.5 years old Not informed Counting, number fact Working memory, Conceptual understanding: additive composition; Conceptual understanding: R2=60.6%
2018 knowledge, arithmetic inhibition, switching, Mathematics achivement: numerical operations of – number line estimation (-0.46);
(IF = 0.25) strategy efficiency and use, visuospatial processing, standardized test (WIAT-II) Procedural understanding: R2=84.8% -
digit recognition, number non-verbal IQ counting (0.25), number fact
line estimation, symbolic knowledge (0.33), working memory
and non-symbolic (0.24).
comparison
Orrantia et al., 52 6.2 years old 1st grade Small (1-9) and large General intellectual ability, Mathematics achievement: Differential and Mathematics achievement: R2=52% –
2018 (6-60) non-symbolic processing speed General Aptitude Battery (BADyG) – calculation two-digit symbolic magnitude
(IF = 0.00) magnitude comparison; and numerical-verbal problems; and a task to comparison (0.26).
one-digit symbolic solve arithmetic problems through mental Mental calculation: R2=54% –
13

magnitude comparison calculation two-digit symbolic magnitude

(1-9) and two-digit comparison (0.26).
symbolic magnitude
comparison (31-79)
Xenidou- 334 5.6 years old Preschool (later, 2nd Counting, magnitude IQ (non-verbal), working General mathematical performance: standardized Early performance (1st grade):
Dervou et al., grade) processing (symbolic and memory (visuospatial test (Cito), ran by schools. Evaluates problem non-verbal IQ (0.15), visuospatial
2018 non-symbolic approximate sketchpad, phonological solving involving measurements, proportions, sketchpad (0,12), central executive
(IF = 0.42) addition, symbolic and loop, and central numbers, numbers and arithmetic relations, (0.12), phonological loop (0.13),

International Journal of Educational Research Open 2–2 (2021) 100035

non-symbolic approximate executive) arithmetic, and complex applications (more than non-symbolic approximate
comparison, symbolic one operation in the same problem) comparison (0.12), symbolic
exact addition) approximate comparison (0.13),
approximate symbolic addition (0.10),
counting (0.23).
Development growth (from 1st to
2nd grade): approximate symbolic
addition (0.26).
Aragón et al., 109 4.9 years old Preschool Symbolic and Verbal working memory, Early mathematics ability – informal thinking: R2 =48,5% – Working memory (0.45),
2019 non-symbolic approximate processing speed and evaluated by TEMA-3 (counting, comparison of processing speed (0.24), vocabulary
(IF = 0.17) comparison, and number vocabulary quantities, informal calculation and basic informal (0.18) and number line estimation.
line estimation. concepts)
Gashaj et al., 136 6.4 years old Preschool (later, 2nd Basic numerical skills: Executive functions: Mathematics achievement: standardized test Basic numerical skills (0.70), updating
2019 grade) magnitude comparison inhibition, switching, based on national curriculum, evaluates (0.27).
(IF = 0.25) and number line updating; and fine motor equations, sequences, addition, and subtraction
estimation skills
(continued on next page)
 C.P. Nogues and B.V. Dorneles
Table 1 (continued)

Study (IF) N Age(average) Evaluated school Specific domain abilities General domain abilities Outcome variables Meaningful predictors and effect size (𝛽)
grades

Gimbert et al., 148 5.7 years old Preschool and 2nd Non-symbolic number Working memory, Mathematics achievement: exact symbolic At preschool: R2 =46% – age (0.22),
2019 and 7.7 years grade comparison, number line vocabulary addition, exact symbolic subtraction, and non-symbolic number comparison
(IF = 0.25) old estimation numerical verbal problems (0.19) e number line estimation
(-0.51).
At 2nd grade: R2 =46% – working
memory (0.44) and number line
estimation (-0.38).
Kiss; Christ, 167 Not informed Preschool and 1st Early numeracy skills: Early Reading skills: Mathematical performance: Group Math Preschool: R2 =37% – early numeracy
2019 grade (Earlymath) evaluates (EarlyReading), evaluates Assessment and Diagnostic Evaluation skills (0.49).
(IF = 0.00) three domains – numbers, pre-literacy elements (GMADE). Standardized test based on the school 1st grade: R2 =61% – early numeracy
relations, and operations (alphabetical principle, curriculum that evaluated language skills(0.65) e early reading skills
decoding, and phonemic understanding and mathematical vocabulary, (0.20).
awareness) ability to identify appropriate representations, to
apply mathematical concepts, and use operation
reasoning to solve problems and calculate
addition and subtraction of whole numbers
Kiss; Nelson; 175 Not informed 1st grade (later, 3rd Early mathematics skills: Mathematical performance: Minnesota Numbers and operations: R2 =41% –
Christ, 2019 grade) decomposing, number Comprehensive Assessment–Series III (MCA-III), decomposing (0.06), number sequence
(IF = 0.17) sequence, numeral standardized test that evaluates numbers and (0.10), verbal subtraction (0.11).
identification (0-120), operations, algebra, geometry and measurements, Algebra: R2 =33% – number sequence
place value, verbal probability, and data analysis (0.10) and computations (0.03)
14

addition, verbal Data interpretation: R2 =25% –

subtraction (oral problems number sequence (0.08).
of one digit), story Geometry and measurements:
problems and R2 =43% – decomposing (0.10).
computations (addition
and subtraction)
Malone et al., 519 5.3 years old Preschool (later, 1st Number knowledge Finger gnosis, executive Arithmetic: evaluated only addition in preschool, Addition evaluated at preschool (0.45),
2019 grade) (identification and writing functions (visual-spatial later evaluated addition and subtraction in the 1st number knowledge (0.29) and

International Journal of Educational Research Open 2–2 (2021) 100035

(IF = 0.17) of numbers), counting, memory, verbal working grade nonsymbolic numerosity
nonsymbolic numerosity memory, inhibition and discrimination (0.22).
discrimination selective attention),
language, and non-verbal
reasoning
Carr et al., 304 7.6 years old 2nd grade (later, 3rd Number sense: place Verbal working memory; Mathematical competence: Criterion Referenced Socioeconomic status (0.11 and 016),
2020 and 4th grades) value, decomposing, spatial ability (mental Competency Test (CRCT) - standardized test that verbal working memory (0.13 and
(IF = 0.00) non-conventional rotation) evaluates ability in calculations, patterns, and 0.09) and spatial abilities (0.18 and
problems and solving of statistic, problem solving and geometry 0.15) – values for students with low
cognitive problems and high performance, respectively.
(mental calculation)
Habermann et 71 4.2 years old Preschool (later, 1st Number identification, Non-verbal IQ, Arithmetic: simple addition (informal test, Arabic numerical knowledge
al., 2020 grade) writing numbers, reading grammatical ability, evaluated in the end of preschool). Evaluated in (identification, reading and number
(IF = 0.00) numbers, counting, vocabulary the 1st grade: sub-test numerical operations of writing) (0.64 – in the arithmetic
symbolic comparison, Wechsler Individual Achievement Test–Second evaluated at preschool and 0.70 at 1st
non-symbolic comparison Edition (WIAT-II) – standardized test, and grade).
arithmetic fluency
C.P. Nogues and B.V. Dorneles International Journal of Educational Research Open 2–2 (2021) 100035

arithmetic performance; and counting and IQ for general mathematical On the abilities of specific domain, those which appear more fre-
performance. quent in the results as explanatory of later mathematical development
It is interesting to note that the regression models considered were were early mathematical abilities, that involved recognition and read-
all those in which researchers inserted all variables at the same time in ing of numbers, discrimination of quantities, knowledge of numerical
the model. However, this is simply an informative report, considering sequence, and here we can include some studies that also understood as
the impossibility of comparing these models and the explanatory values early abilities the knowledge of counting and magnitudes comparison. It
found, as this will depend heavily on the chosen abilities, as well as the is important to mention that, except for one study, all studies which con-
tasks used, and the sample evaluated. Nonetheless, it is important to sidered early mathematical abilities in their analysis found significant
highlight that some results, regardless of these various factors, are sim- results in their explanations for later mathematical performance.
ilar, even in different countries and cultures. This allows us to raise the Other important factor to be highlighted is that these abilities can
hypothesis that some cognitive abilities might be considered universal have a higher or lower value of explanation depending on the school
to the development of mathematical knowledge. grade the child is and what is considered as mathematical performance.
On Table 1 we present a synthesis of all the studies analyzed in this From the results of this systematic review, we can perceive that abil-
review organized by chronological order. We also calculated the impact ities such as the approximate number system (ANS), the recognition
factor (IF) of each article in this review, that is, the studies with the and reading of numbers, and the identification of quantities are more
highest number of citations. To do so, we considered the total number explicative of the performance in children starting schooling, that is,
of citations of each article divided by the time (in months) since its Preschool, 1st , and 2nd grade. At the end of the Primary school, i.e., 3rd ,
publication until the date we research it. The number of citations was 4th , and 5th grades, other more complex abilities appear as prediction
provided by searching the title of each article in the site Google Scholar, values, which involve arithmetic knowledges, such as calculations with
a widely used search tool that can identify the number of works in which mathematical operations and arithmetic facts.
each article was cited. However, we know that some factors can interfere Considering the type of mathematical achievement evaluated, we
in this calculation, such as the language in which the article was written can highlight that the performances focused on arithmetic and on gen-
and the time of publication, since some of them were published in 2020, eral mathematics depend on working memory and early numerical abil-
the same year of this review. The studies with the highest impact factor ities so that students can have a better mathematical development. The
were: Bull et al., 2008 (IF=12.22); Jordan et al., 2009 (IF=6.97); and results also indicate some specificities, as the comparison of magnitudes
Geary, 2011 (IF=6.68). From that, we understand that these articles and the abilities of reading and writing as explicative of arithmetic per-
were the most cited ones up to the moment they were included in the formance, counting, and intelligence as predictors of general mathe-
review, therefore, they have more impact in the academic area and are matical achievement. However, some abilities are recurrent regardless
the most consulted ones by researchers in the topic. of school grade or mathematical task evaluated, as is the case of in-
telligence, working memory, counting, number estimation, and early
numerical abilities, that include number recognition (understanding of
4. Discussion number system), comparison of symbolic quantities and number sense.
In addition, another aspect that is not considered in the analyzes
The aim of this systematic review was to identify the cognitive abil- and that arises as important to be discussed is the diversity of these
ities of general and specific domain indicated in the literature as pre- samples. Even controlling for age, intelligence and socioeconomic fac-
dictors of mathematical performance in students from primary school. tors, these studies were conducted in different countries or in different
Though the search was done only with an upper limit for the publica- regions within the same country, and social, cultural and even spoken
tion date, i.e., it could include works done in any year until the time of language aspects are also involved in the teaching and learning process,
the search, most included studies (69.35%) were published in the last which can be underlying mathematical performance. Even so, some cog-
decade, between 2011 and 2020, and most of them in the United States. nitive skills proved to be significant predictors regardless of these fac-
This recent increase in the literature related to this theme can be due tors, which allows us to think about the possibility of universal cogni-
to the emergence of more research projects connected to mathematical tive skills for the development of early numerical knowledge. Therefore,
education and numerical cognition in the last years, focusing especially such skills could be inserted in the school context as tasks to stimulate
on the development of numerical knowledge, mainly in the first years the development of mathematical knowledge and prevent possible later
of school. From this synthesis of investigated studies, we could see that mathematical difficulties.
most part consisted of longitudinal researches, whose initial evaluations Besides this, some limitations of this literature systematic review
were done with children from Preschool, that is, even before the system- should be taken into consideration. First, we did not include compar-
atized teaching of mathematics. Besides this, we also found that there ative studies between different countries, which could control cultural
was a concern in evaluating abilities from general and specific domains differences and language. This allowed the direct analysis of checking
in these studies, what made it possible to identify similarities and some if, in different cultures, the predictor abilities are the same, thus limit-
inconsistences in the findings. ing the generalization of the results. Secondly, most studies included in
Therefore, from the results, we can highlight that the cognitive abil- this synthesis focused on the evaluation of Preschool students, restrict-
ity from the general domain most frequently pointed as predictor of ing the generalization of the findings and the knowledge base on cog-
mathematical performance was working memory, regardless the school nitive abilities through the school grades. Besides this, as more studies
grade or mathematical task evaluated as outcome. However, it is worth focus on the evaluation of cognitive abilities in the first years of school-
highlighting that some studies that measured working memory did not ing, i.e., Preschool and 1st grade, there were more predictors related to
find that it was a significant variable to explain mathematical perfor- early numerical abilities because participants’ schooling level and age
mance. In the attempt to explain this data, we can take into considera- do not allow the evaluation of more complex abilities. Third and last,
tion some factors, such as the instruments used for the evaluation and the differences of theories and definitions on what the authors consider
other abilities of the specific domain that were inserted in the regression as early numerical abilities, including the differences on the statistical
model, considering that numerical abilities tend to overlap themselves in analysis, allow different results and interpretations, what also limits a
this direct relation between working memory and mathematical perfor- more consistent generalized understanding on the abilities that explain
mance. When this overlapping takes place, working memory appears as mathematical performance.
a mediating ability, i.e., having an indirect effect in the relation between Another important aspect to be highlighted is the difficulty to artic-
numerical abilities and mathematical performance (Gimbert et al., 2019; ulate research in cognitive psychology and educational theories, more
Passolunghi & Lanfranchi, 2012). than that with educational practice. Even though research in the field of

15
C.P. Nogues and B.V. Dorneles International Journal of Educational Research Open 2–2 (2021) 100035

numerical cognition has been on the rise in recent years, as evidenced Acknowledgments
in this systematic review, the challenge lies in transferring the results
of this research to teaching practice (Simplicio et al., 2020). Studies in This study was financed in part by the Coordenação de Aperfeiçoa-
cognitive psychology, generally because they follow consistent research mento de Pessoal de Nível Superior - Brasil (CAPES) – Finance code 001.
methods, end up disregarding environmental variables and teaching re-
quirements. Therefore, the results found may not be enough to derive
References
effective and direct suggestions for educational practice (Simplicio et al.,
2020). However, research in this area provides new perspectives on Andersson, U. (2008). Working memory as a predictor of written arithmetical skills in
teaching and learning, and some indications that are consistent with stu- children: The importance of central executive functions. British Journal of Educational
Psychology, 78(2), 181–203. 10.1348/000709907X209854.
dents’ learning needs and new tools for more effective teaching. Never-
Aragón, E., Cerda, G., Delgado, C., Aguilar, M., & Navarro, J. I. (2019). Individual dif-
theless, in order for these research results to have concrete applicability ferences in general and specific cognitive precursors in early mathematical learning.
at school, more studies inspired by interdisciplinary use are needed to Psicothema, 31(2), 156–162. 10.7334/psicothema2018.306.
guide this transfer and enable this approximation between theory and Aunio, P., & Niemivirta, M. (2010). Predicting children’s mathematical performance in
grade one by early numeracy. Learning and Individual Differences, 20(5), 427–435.
practice (Simplicio et al., 2020). Aunola, K., Leskinen, E., Lerkkanen, M. K., & Nurmi, J. E. (2004). Developmental dynamics
Besides that, according to the results and limitations of the present of math performance from preschool to grade 2. Journal of Educational Psychology,
study, a possibility for future research is to consider the development of 96(4), 699–713. 10.1037/0022-0663.96.4.699.
Bartelet, D., Vaessen, A., Blomert, L., & Ansari, D. (2014). What basic number
mathematical knowledge through time, i.e., that evaluates the predictor processing measures in kindergarten explain unique variability in first-grade
cognitive abilities in other grades of schooling, in addition to Primary arithmetic proficiency? Journal of Experimental Child Psychology, 117(1), 12–28.
Education. Thus, more evidence can be added in the area, including 10.1016/j.jecp.2013.08.010.
Brasil. (2018). Base Nacional comum (BNCC). Mec, 600.
more complex mathematical abilities. Besides this, we suggest the inclu- 10.1017/CBO9781107415324.004.
sion of participants from other cultures and regions as a way to replicate Bull, R., Espy, K. A., & Wiebe, S. A. (2008). Short-term memory, working memory,
and extend the results of current research. and executive functioning in preschoolers: Longitudinal predictors of mathemat-
ical achievement at age 7 years. Developmental Neuropsychology, 33(3), 205–228.
Furthermore, the findings of this systematic review provide a base
10.1080/87565640801982312.
for the scientific community in the area as well as for the pedagogi- Butterworth, B. (2005). The development of arithmetical abilities. Jour-
cal practice and related professionals, since it contributes to understand nal of Child Psychology and Psychiatry and Allied Disciplines, 46(1), 3–18.
10.1111/j.1469-7610.2004.00374.x.
and evaluate key cognitive abilities that support long-term mathemati-
Carr, M., Horan, E., Alexeev, N., Barned, N., Wang, L., & Otumfuor, B. (2020). A longi-
cal learning. More than that, the results gather evidence that children’s tudinal study of spatial skills and number sense development in elementary school
performance in early numerical abilities and working memory is deter- children. Journal of Educational Psychology, 112(1), 53–69. 10.1037/edu0000363.
minant to learn mathematics. Such abilities can be easily evaluated by Casey, B. M., Lombardi, C. M. P., Pollock, A., Fineman, B., & Pezaris, E. (2017). Girls’
spatial skills and arithmetic strategies in first grade as predictors of fifth-grade
teachers and, moreover, are liable to be stimulated even before formal analytical math reasoning. Journal of Cognition and Development, 18(5), 530–555.
schooling and can last as a work possibility in the classroom through 10.1080/15248372.2017.1363044.
the first years of schooling. School incentive in the development of these Ching, B. H.-H., & Nunes, T. (2017). The importance of additive reasoning in children’s
mathematical achievement: A longitudinal study. Journal of Educational Psychology,
abilities will entail not only an improvement on students’ mathematical 109(4), 477–508 VolIssue.
development, but also act as a resource to attenuate the already-existing Chu, F. W., vanMarle, K., & Geary, D. C. (2015). Early numerical foundations of young
difficulties and prevent later mathematical difficulties. children’s mathematical development. Journal of Experimental Child Psychology, 132,
205–212. 10.1016/j.jecp.2015.01.006.
Cirino, P. T. (2011). The interrelationships of mathematical precursors in
kindergarten. Journal of Experimental Child Psychology, 108(4), 713–733.
10.1016/j.jecp.2010.11.004.
5. Conclusion De Smedt, B., Janssen, R., Bouwens, K., Verschaffel, L., Boets, B., & Ghesquière, P. (2009).
Working memory and individual differences in mathematics achievement: A longitu-
dinal study from first grade to second grade. Journal of Experimental Child Psychology,
We revised studies that evaluated cognitive abilities as precursors
103(2), 186–201. 10.1016/j.jecp.2009.01.004.
of later mathematical performance intending to verify the abilities that De Smedt, B., Verschaffel, L., & Ghesquière, P. (2009). The predictive value of numerical
are more commonly measured and the frequency of identification. In all magnitude comparison for individual differences in mathematics achievement. Jour-
nal of Experimental Child Psychology, 103(4), 469–479. 10.1016/j.jecp.2009.01.010.
schooling levels and considering all types of mathematical performance
Desoete, A., Stock, P., Schepens, A., Baeyens, D., & Roeyers, H. (2009). Classification, se-
evaluated, early numerical abilities and working memory were the cog- riation, and counting in grades 1, 2, and 3 as two-year longitudinal predictors for low
nitive abilities that appeared more frequently as having a significant achieving in numerical facility and arithmetical achievement? Journal of Psychoedu-
explanation value for later mathematical performance. Besides this, we cational Assessment, 27(3), 252–264. 10.1177/0734282908330588.
Fanari, R., Meloni, C., & Massidda, D. (2018). Visuospatial working memory and early
highlight that regardless the qualitative aspects of the sample, as coun- math skills in first grade children. In Proceedings of the 15th international conference on
try of origin, main language, social and cultural differences, these same cognition and exploratory learning in the digital age, CELDA 2018, Celda (pp. 127–133).
abilities appear as significant predictors of mathematical performance. Foster, M. E., Anthony, J. L., Clements, D. H., & Sarama, J. H. (2015). Processes in the
development of mathematics in kindergarten children from Title 1 schools. Journal of
In this sense, we raise the hypothesis that such cognitive abilities might Experimental Child Psychology, 140, 56–73. 10.1016/j.jecp.2015.07.004.
be universal to the development of mathematical knowledge in students Fuchs, L. S., Fuchs, D., Compton, D. L., Powell, S. R., Seethaler, P. M., Capizzi, A. M.,
in Primary school. Therefore, there is the need to emphasize in school Schatschneider, C., & Fletcher, J. M. (2006). The cognitive correlates of third-grade
skill in arithmetic, algorithmic computation, and arithmetic word problems. Journal
the tasks that stimulate those abilities, to prevent and remediate math- of Educational Psychology, 98(1), 29–43. 10.1037/0022-0663.98.1.29.
ematical difficulties. More than this, the results point ways to future Fuchs, L. S., Geary, D. C., Fuchs, D., Compton, D. L., & Hamlett, C. L. (2014). Sources of
researches, indicating the importance of a longitudinal follow-up of stu- individual differences in emerging competence with numeration understanding ver-
sus multidigit calculation skill. Journal of Educational Psychology, 106(2), 482–498
dents’ mathematical performance through schooling years, even in more
VolIssue.
advanced grades, and consider the inclusion of participants of different Fuhs, M. W., Hornburg, C. B., & McNeil, N. M. (2016). Specific early number skills mediate
countries and cultures to extend the results of current researches. the association between executive functioning skills and mathematics achievement.
Developmental Psychology, 52(8), 1217–1235. 10.1037/dev0000145.
Gashaj, V., Oberer, N., Mast, F. W., & Roebers, C. M. (2019). The relation between ex-
ecutive functions, fine motor skills, and basic numerical skills and their relevance
for later mathematics achievement. Early Education and Development, 30(7), 913–926.
Declarations of Competing Interest 10.1080/10409289.2018.1539556.
Geary, D. C. (2011). Cognitive predictors of achievement growth in mathematics: A 5-year
The authors declare that they have no known competing financial longitudinal study. Developmental Psychology, 47(6), 1539–1552. 10.1037/a0025510.
Geary, D. C., & van Marle, K. (2016). Young children’s core symbolic and nonsymbolic
interests or personal relationships that could have appeared to influence quantitative knowledge in the prediction of later mathematics achievement. Develop-
the work reported in this paper. mental Psychology, 52(12), 2130–2144. 10.1037/dev0000214.

16
C.P. Nogues and B.V. Dorneles International Journal of Educational Research Open 2–2 (2021) 100035

Gilligan, K. A., Flouri, E., & Farran, E. K. (2017). The contribution of spatial ability to Nunes, T., & Bryant, P. (1997). Crianças fazendo matemática Artmed.
mathematics achievement in middle childhood. Journal of Experimental Child Psychol- Nunes, T., Bryant, P., Evans, D., Bell, D., Gardner, S., Gardner, A., & Carra-
ogy, 163, 107–125. 10.1016/j.jecp.2017.04.016. her, J. (2007). The contribution of logical reasoning to the learning of mathemat-
Gilmore, C., Clayton, S., Cragg, L., McKeaveney, C., Simms, V., & Johnson, S. (2018). ics in primary school. British Journal of Developmental Psychology, 25(1), 147–166.
Understanding arithmetic concepts: The role of domain-specific and domain-general 10.1348/026151006X153127.
skills. PLoS ONE, 13(9), 1–20. 10.1371/journal.pone.0201724. Nyroos, M., & Wiklund-Hornqvist, C. (2012). The association between working mem-
Gilmore, C. K., McCarthy, S. E., & Spelke, E. S. (2010). Non-symbolic arithmetic abilities ory and educational attainment as measured in different mathematical subtopics in
and mathematics achievement in the first year of formal schooling. Cognition, 115(3), the swedish national assessment: Primary education. Educational Psychology, 32(2),
394–406. 10.1016/j.cognition.2010.02.002. 239–256.
Gimbert, F., Camos, V., Gentaz, E., & Mazens, K. (2019). What predicts mathematics Orrantia, J., San Romualdo, S., Matilla, L., Sánchez, R., Múñez, D., & Verschaffel, L. (2017).
achievement? Developmental change in 5- and 7-year-old children. Journal of Ex- Marcadores nucleares de la competencia aritmética en preescolares. Psychology, Soci-
perimental Child Psychology, 178, 104–120. 10.1016/j.jecp.2018.09.013. ety, & Education, 9(1), 121–124.
Habermann, S., Donlan, C., Göbel, S. M., & Hulme, C. (2020). The critical role of Arabic nu- Orrantia, J., San Romualdo, S., Sánchez, R., Matilla, L., Muñez, D., & Verschaffel, L. (2018).
meral knowledge as a longitudinal predictor of arithmetic development. Journal of Ex- Procesamiento de magnitudes numéricas y ejecución matemática. Revista de Edu-
perimental Child Psychology, 193(2020), Article 104794. 10.1016/j.jecp.2019.104794. cación, 381, 133–154. 10.4438/1988-592X-RE-2017-381-383.
Hawes, Z., Moss, J., Caswell, B., Seo, J., & Ansari, D. (2019). Relations between Passolunghi, M. C., Vercelloni, B., & Schadee, H. (2007). The precursors of mathematics
numerical, spatial, and executive function skills and mathematics achievement: learning: Working memory, phonological ability and numerical competence. Cognitive
A latent-variable approach. Cognitive Psychology, 109(December 2018), 68–90. Development, 22(2), 165–184. 10.1016/j.cogdev.2006.09.001.
10.1016/j.cogpsych.2018.12.002. Passolunghi, M. C., & Lanfranchi, S. (2012). Domain-specific and domain-general
Hecht, S. A., Torgesen, J. K., Wagner, R. K., & Rashotte, C. A. (2001). The relations between precursors of mathematical achievement: A longitudinal study from kinder-
phonological processing abilities and emerging individual differences in mathemat- garten to first grade. British Journal of Educational Psychology, 82(1), 42–63.
ical computation skills: A longitudinal study from second to fifth grades. Journal of 10.1111/j.2044-8279.2011.02039.x.
Experimental Child Psychology, 79(2), 192–227. 10.1006/jecp.2000.2586. Purpura, D. J., & Logan, J. A. R. (2015). The nonlinear relations of the approximate num-
Jordan, N. C., Glutting, J., & Ramineni, C. (2010). The importance of number sense to ber system and mathematical language to early mathematics development. Develop-
mathematics achievement in first and third grades. Learning and Individual Differences, mental Psychology, 51(12), 1717–1724.
20(2), 82–88. Reigosa-Crespo, V., González-Alemañy, E., León, T., Torres, R., Mosquera, R., & Valdés-
Jordan, N. C., Kaplan, D., Locuniak, M. N., & Ramineni, C. (2007). Predicting first-grade Sosa, M. (2013). Numerical capacities as domain-specific predictors beyond early
math achievement from developmental number sense trajectories. Learning Disabilities mathematics learning: A longitudinal study. PLoS ONE, 8(11), 1–11. 10.1371/jour-
Research & Practice, 22(1), 36–46. 10.1111/j.1540-5826.2007.00229.x. nal.pone.0079711.
Jordan, N. C., Kaplan, D., Ramineni, C., & Locuniak, M. N. (2009). Early math matters: Sánchez, R., Matilla, L., & Orrantia, J. (2017). Relaciones entre procesamiento fonológico y
kindergarten number competence and later mathematics outcomes. Developmental diferencias individuales en ejecución matemática: Un estudio longitudinal. II Congreso
Psychology, 45(3), 850–867. 10.1038/jid.2014.371. Internacional Virtual Sobre La Educación En El Siglo XXI, marzo, 432–442.
Kiss, A. J., & Christ, T. J. (2019). Screening for math in early grades: Is reading enough? Sasanguie, D., Göbel, S. M., Moll, K., Smets, K., & Reynvoet, B. (2013). Approximate num-
Assessment for Effective Intervention, 45(1), 38–50. 10.1177/1534508418766410. ber sense, symbolic number processing, or number-space mappings: What underlies
Kiss, A. J., Nelson, G., & Christ, T. J. (2019). Predicting third-grade mathematics achieve- mathematics achievement? Journal of Experimental Child Psychology, 114(3), 418–431.
ment: A longitudinal investigation of the role of early numeracy skills. Learning Dis- 10.1016/j.jecp.2012.10.012.
ability Quarterly, 42(3), 161–174. 10.1177/0731948718823083. Sasanguie, D., Van den Bussche, E., & Reynvoet, B. (2012). Predictors for mathematics
Krajewski, K., & Schneider, W. (2009). Early development of quantity to number-word achievement? Evidence from a longitudinal study. Mind, Brain, and Education, 6(3),
linkage as a precursor of mathematical school achievement and mathematical diffi- 119–128.
culties: Findings from a four-year longitudinal study. Learning and Instruction, 19(6), Simmons, F., Singleton, C., & Horne, J. (2008). Brief report - Phonological awareness and
513–526. 10.1016/j.learninstruc.2008.10.002. visual-spatial sketchpad functioning predict early arithmetic attainment: Evidence
LeFevre, J.-A., Fast, L., Skwarchuk, S.-L., Smith-Chant, B. L., Bisanz, J., Kamawar, D., & from a longitudinal study. European Journal of Cognitive Psychology, 20(4), 711–722.
Penner-Wilger, M. (2010). Pathways to mathematics: Longitudinal predictors of per- 10.1080/09541440701614922.
formance. Child Development, 81(6), 1753–1767. 10.1111/j.1467-8624.2010.01508.x. Simplicio, H., Gasteiger, H., Dorneles, B. V., Grimes, K. R., Haase, V. G., Ruiz, C.,
Libertus, M. E., Feigenson, L., & Halberda, J. (2011). Preschool acuity of the approxi- Liedtke, F. V., & Moeller, K. (2020). Cognitive research and mathematics educa-
mate number system correlates with school math ability. Developmental Science, 14(6), tion—how can basic research reach the classroom? Frontiers in Psychology, 11(April),
1292–1300. 10.1038/jid.2014.371. 1–5. 10.3389/fpsyg.2020.00773.
Libertus, M. E., Feigenson, L., & Halberda, J. (2013). Is approximate number precision Szucs, D., Devine, A., Soltesz, F., Nobes, A., & Gabriel, F. (2014). Cognitive components
a stable predictor of math ability? Learning and Individual Differences, 25, 126–133. of a mathematical processing network in 9-year-old children. Developmental Science,
10.1016/j.lindif.2013.02.001. 17(4), 506–524.
Locuniak, M. N., & Jordan, N. C. (2008). Using kindergarten number sense to predict Tobia, V., Bonifacci, P., & Marzocchi, G. M. (2016). Concurrent and longitudinal predictors
calculation fluency in second grade. Journal of Learning Analytics, 41(5), 451–459. of calculation skills in preschoolers. European Journal of Psychology of Education, 31(2),
10.1177/0022219408321126.Using. 155–174.
Malone, S. A., Burgoyne, K., & Hulme, C. (2019). Number knowledge and the approximate Toll, S. W. M., Kroesbergen, E. H., & Van Luit, J. E. H. (2016). Visual working memory and
number system are two critical foundations for early arithmetic development. Journal number sense: Testing the double deficit hypothesis in mathematics. British Journal of
of Educational Psychology. 10.1037/edu0000426. Educational Psychology, 86(3), 429–445. 10.1111/bjep.12116.
Martin, R. B., Cirino, P. T., Sharp, C., & Barnes, M. (2014). Number and counting skills Van der Ven, S. H. G., Kroesbergen, E. H., Boom, J., & Leseman, P. P. M. (2012).
in kindergarten as predictors of grade 1 mathematical skills. Learning and Individual The development of executive functions and early mathematics: A dy-
Differences, 34, 12–23. 10.1016/j.lindif.2014.05.006. namic relationship. British Journal of Educational Psychology, 82(1), 100–119.
Mazzocco, M. M. M., Feigenson, L., & Halberda, J. (2011). Preschoolers’ precision of the 10.1111/j.2044-8279.2011.02035.x.
approximate number system predicts later school mathematics performance. PLoS vanMarle, K., Chu, F. W., Li, Y., & Geary, D. C. (2014). Acuity of the approximate num-
ONE, 6(9). 10.1371/journal.pone.0023749. ber system and preschoolers’ quantitative development. Developmental Science, 17(4),
Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for 492–505.
systematic reviews and meta-analyses: The PRISMA statement. PLoS Med, 6(7), Article Wong, T. T.-Y., & Ho, C. S.-H. (2017). Component processes in arithmetic word-problem
E1000097. 10.1371/journal.pmed1000097. solving and their correlates. Journal of Educational Psychology, 109(4), 520–531.
Moher, D., Shamseer, L., Clarke, M., et al. (2015). Preferred reporting items for system- Xenidou-Dervou, I., Van Luit, J. E. H., Kroesbergen, E. H., Friso-van den Bos, I.,
atic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev, 4(1). Jonkman, L. M., van der Schoot, M., & van Lieshout, E. C. D. M. (2018). Cognitive
10.1186/2046-4053-4-1. predictors of children’s development in mathematics achievement: A latent growth
Nguyen, T., Watts, T. W., Duncan, G. J., Clements, D. H., Sarama, J. S., Wolfe, C., & modeling approach. Developmental Science, 21(6), 1–14. 10.1111/desc.12671.
Spitler, M. E. (2016). Which preschool mathematics competencies are most predictive
of fifth grade achievement? Early Childhood Research Quarterly, 36(2016), 550–560.
10.1016/j.ecresq.2016.02.003.Which.

17