1

British Journal of Developmental Psychology (2017)

© 2017 The British Psychological Society
www.wileyonlinelibrary.com

Assessment and conceptual remediation of basic

calculation skills in elementary school students
Elen S. Karakonstantaki1* , Panagiotis G. Simos2,
Vamvukas Michalis3 and Sifis Micheloyannis4
1
School of Medicine, University of Crete, Heraklion, Greece
2
Division of Psychiatry, Department of Medicine, School of Medicine, University of
Crete, Heraklion, Greece
3
Department of Pedagogy, University of Crete, Rethymnon, Greece
4
Medical Department (Labor Widen), School of Medicine, University of Crete,
Heraklion, Greece
The specific domain model for math disabilities postulates a core number deficit which
presents a prime target for remedial interventions. This longitudinal study identified two
groups of Grade 3 students based on their basic calculation abilities: students with
persistent difficulties through Grade 4 (PD group) and students whose performance
improved into the average range (IP group). Baseline data revealed a distinct cognitive
profile for students in the PD group featuring predominant deficits in symbolic number
processing. A conceptual intervention based on explicit teaching of basic arithmetic
procedures was implemented when students attended Grade 5 or 6. Students in the PD
group benefited more from the programme, especially in performing written calculations
and in multiplication speed.

Statement of contribution
What is already known on this subject?
 Most interventions focus on young students’ basic arithmetical skills to prevent serious math
problems in future.
 Few interventions target older students who often face persistent math difficulties.
 These interventions are usually procedural and focus on age-appropriate math skills.

What does this study add?

 A conceptual intervention was implemented to remediate basic calculation deficits at the end of
elementary school.
 The aim was to help students compensate for their gaps in knowledge and motivate them to engage
in math activities.
 Neuropsychological testing of arithmetic abilities revealed difficulties in symbolic number
processing.

Math education is receiving increased attention in recent years due to the critical role of
math skills in technology research and industry bearing significant implications for
economic growth and individual success in a highly competitive workplace. It is

*Correspondence should be addressed to Elen S. Karakonstantaki, School of Medicine, University of Crete, G. Papandreou 33 &
Xrisostomou, Heraklion 71306, Crete, Greece (email: [email protected]).

DOI:10.1111/bjdp.12214
2 Elen S. Karakonstantaki et al.

estimated that developmental dyscalculia, a disorder associated with an impairment in

understanding and manipulating numerosity, which manifests as severe difficulty in basic
math skills, affects about 5% of primary school students (Butterworth, 2010; Kaufmann &
von Aster, 2012). Difficulties in math may also derive from inadequate teaching methods
or poor motivation, may be associated with other learning disabilities, and often persist
through adulthood (Miller & Mercer, 1998). Certain common characteristics of math
curricula may contribute to the phenomenon by focusing on the transfer of formal
knowledge. These teaching approaches tend to disregard the intuitive mathematical
knowledge children develop from real-life experiences even before entering school
(Carraher, Carraher, & Schliemann, 1985) and the diversity of developmental trajectories
in learning math (Scherer, Beswick, DeBlois, Healy, & Moser Opitz, 2016). Diagnostic
classification of difficulties in math is challenging as development of arithmetic skills is
heterogeneous following distinct trajectories that vary across children and contexts. As a
result of person- and context-related factors, individual arithmetic performance may
display significant variability over time. The heterogeneity of factors causally related to
math learning difficulties contributes to the lack of uniform performance criteria in both
research and clinical settings (Kaufmann et al., 2013).

Cognitive underpinnings of math skills and math disability

Math difficulties are often paralleled by inadequate development of various cognitive
abilities including language, memory, or executive functioning (Geary & Hoard, 2005;
Szucs, Devine, Soltesz, Nobes, & Gabriel, 2013). A core deficit in number sense has
been lately hypothesized for students with severe math impairment. Number sense
refers to the ability to represent and manipulate numerical quantities on a mental-
analogue of a number line, or as an array of objects. Number sense deficits lead to
impaired understanding of numbers and their relations, affecting performance on a
variety of tasks requiring semantic manipulation of numbers (presented as either
Arabic digits, numberwords, or arrays of objects). It is involved in subitizing (quick
estimation of quantity), approximate calculation, comparison, bisection (finding which
number lies between two others), and computations, such as subtraction, which
cannot be solved based on knowledge (and recall) of basic math facts. Conversely,
exact calculation requires qualitatively different skills, which rely heavily either on
language capacities, as in the case of simple addition and multiplication where math
fact retrieval is necessary, or on complex visuospatial representation and manipulation
capacities, as in the case of multidigit operations which are handled as strings of
Arabic numerals (Dehaene, Piazza, Pinel, & Cohen, 2003). Many studies have
suggested a deficit in understanding numerosity as the more common cause of
difficulties in basic number skills involving quantity processing (e.g., counting and
comparison; Butterworth, 2005, 2010; Butterworth, Varma, & Laurillard, 2011;
Landerl, Bevan, & Butterworth, 2004; Mussolin, Mejias, & Noёl, 2010). Other
researchers have emphasized deficiencies in symbolic tasks (involving Arabic digits
or numberwords) in the presence of intact capacity to perform non-symbolic tasks
(e.g., mental manipulation of dots) and have proposed a dissociation between
symbolic and non-symbolic quantity representation and processing as the most
common cause of math learning difficulties. According to this view, numerosity may
be intact, but poor math achievement is caused by a deficit in accessing numerical
meaning through symbols (De Smedt & Gilmore, 2011; Geary, Hoard, Nugent, &
Bailey, 2012; Rousselle & Noёl, 2007; Wilson & Dehaene, 2007).
 Evaluation and remediation of calculation skills 3

Interventions for math disabilities

In a recent meta-analysis of math intervention studies, Chodura, Kuhn, and Holling (2015)
have highlighted employment of direct or assisted instructional procedures, fostering
basic arithmetic competencies and the individualized delivery of instruction as effective
intervention practices. The age of participating students and the intensity of interventions
were not found to correlate with intervention effectiveness. Interestingly, even the use of
computers was not found to enhance significantly students’ performance although
computer-assisted interventions for math disabilities have been considered useful by
others (see Rӓsӓnen, Salminen, Wilson, Aunio, & Dehaene, 2009 for a review). The
majority of available math teaching programmes combine a focus on procedural skills
with explicit instruction methods. For instance, the Connecting Math Concepts
programme (Engelmann, Carnine, Kelly, & Engelmann, 1991–1995) focuses on explicit
instruction of basic math concepts introduced in a systematic way, emphasizing links
between different aspects of mathematic knowledge while providing guided practice.
Cover, Copy and Compare (Lee & Tingstron, 1994) involves presenting students with the
solution of a math problem, which is then taken away, and students are asked to
reproduce it and compare their copy to the initially presented solution. Math Flash was
developed to enhance automatic retrieval of number combinations (Fuchs et al., 2009).
Although manipulatives, number lines, and finger counting are used, the main focus lies
on learning strategies for fast and accurate calculation of number facts through extended
drill and practice and technological support.
Whereas procedural approaches involve systematic, step-by-step teaching of math
procedures and focus on how to perform operations and solve arithmetic problems,
conceptual approaches aim to improve understanding of the math principles underlying
arithmetic operations and stress the importance of knowing why we perform a certain
procedure. Conceptual interventions, targeting quantity representation and manipula-
tion, have been used in school settings for elementary students with typical math
performance (Hyde, Khanum, & Spelke, 2014) and students with developmental
dyscalculia (Kaufmann, Handl, & Th€ ony, 2003) with encouraging results. Mathematics
Recovery (Wright, 2003) is an adaptive tutoring intervention, differentiating instruction
and goals according to the developmental stage of the student. The Numeracy Recovery
Programme (Dowker, 2001) focuses on the specific components of numeracy where
students encounter difficulties (e.g., counting principles, word problem-solving, trans-
lation between concrete, verbal and/or numerical formats of problems). Based on
Numeracy Recovery, Catch Up Numeracy has been recently developed (Holmes &
Dowker, 2013) as a non-intensive, group-based remedial programme for students with
low attainment in numeracy tasks that persist in time. Number Worlds (Griffin, 2004) was
designed to enhance number sense through establishing connections, allowing active
exploration and encouraging discussion of concepts in a social context.
In spite of recommendations that effective interventions should target weaknesses in
specific components of numeracy and should be differentiated according to the
manifested impairments (Chan, Au, & Tang, 2013; Dowker & Sigley, 2010), remediation
programmes often focus on helping students with learning disabilities catch up with their
peers and be able to follow the class curriculum, ignoring any underlying deficits
(Butterworth et al., 2011; Kaufmann & von Aster, 2012).
Moreover, the majority of widely available intervention/prevention programmes
target students in the early grades (K-3), while older students, often through secondary
education, also face difficulties in basic computation and numeration, in understanding
complex concepts, such as decimals or measurement, and in applying math concepts
4 Elen S. Karakonstantaki et al.

(Jones, Wilson, & Bhojwani, 1998). Data on the beneficial effects of conceptual
interventions for older students have been reported. For instance, Ketterlin-Geller, Chard,
and Fien (2008) attempted to reteach fundamental math concepts and principles to low-
math-achieving fifth graders and observed notable improvement on a standardized math
screener as compared to a treatment as usual, control group. The Every Day Counts
programme was administered over a sixth-month period to fifth graders teaching basic
concepts regarding the calendar and everyday student routines demonstrating modest but
significant results (Great Source, 2006). Another conceptual approach using a concrete-
to-representational-to-abstract sequence of instruction for improving algebra skills of
middle school students with math difficulties benefited students to the same degree as
those with average and above-average math achievement (Witzel, 2005).

The present study

The present study assessed the value of MyMath intervention, which was inspired by the
previous work and was developed to meet the needs of students experiencing persistent
math difficulties in advanced elementary grades. MyMath focuses on conceptual
understanding of number processing strategies and aims to promote addition and
subtraction skills in older elementary school students (mean age 11.2  0.5 years). Due
to the lack of standardized norms for math ability in the Greek education system, we
adopted the criterion of persistence of math difficulties over time as proposed by
Mazzocco and Myers (2003). Assessing math performance at two different time points is
expected to increase classification reliability as Silver, Pennett, Black, Fair, and Balise
(1999) found that approximately half of students with an arithmetic disability diagnosis at
one point in time did not display deficits in math tests 19 months later. The study
examined the performance of two intervention groups: a group of students whose math
difficulties persisted over time (Persistent Difficulty; PD) and a group whose math
performance improved (Improved Performance; IP). We hypothesized that the two
groups of students facing calculation difficulties would have distinct cognitive profiles,
with those facing persistent problems showing predominant number sense deficits. We
also hypothesized that enhancement of conceptual knowledge would help students in
both groups automatize the retrieval and application of basic number facts and
procedures leading to improved calculation performance.

Method
Participants
With permission from the University of Crete, School of Medicine Ethics Committee (in
agreement with the Declaration of Helsinki) and the Research Department of the Greek
Institute of Education, and after parents had signed an informed consent, basic calculation
skills screening was conducted in 48 elementary schools, chosen so as to represent the
population of the Prefecture of Heraklion (32 from urban, 7 from small towns, 9 from
rural areas). A total of 1,382 students (723 boys, 659 girls) attending Grade 3 (Time 1) were
administered the short, 13-item version of the Greek Math Screening Test (GMST-13).
Two groups of students were selected based on their scores on GMST-13: a low-
achieving group of students scoring below the 16th percentile (n = 122) and a typical
performance (TP) group of students scoring above the 35th percentile (n = 122). Parental
consent for further cognitive assessment was obtained for 115 students (67 boys, 48 girls)
who were administered an additional cognitive and comprehensive math assessment at
 Evaluation and remediation of calculation skills 5

the beginning of Grade 4 (Time 2). Sixty-seven (36 low-performing and 31 typical) did not
meet any of the following exclusionary criteria: (1) estimated Full Scale IQ < 80 (average
Vocabulary and Block Design scaled scores < 6), (2) age-adjusted score below the 25th
percentile on the average of standardized word and pseudoword reading accuracy tasks,
(3) an age-adjusted T score higher than 70 points on any of the eight syndrome Child
Behavior Checklist scales, and (4) students born in immigrant families who did not attend
age-appropriate class or were not fluent in Greek language as indicated by their teacher.
None of the students had repeated a grade. These students were re-administered the
GMST-13 in Grade 4 (Time 3) and based on their performance were assigned to two
groups: (1) Students with scores < 50th percentile were considered to exhibit persistent
calculation difficulties (PD, 19 students, 10 boys, 9 girls, mean age 9.77  0.4 years) and
(2) students performing >50th percentile were considered to exhibit improved
calculation performance (IP, 17 students, 9 boys, 8 girls, mean age 9.82  0.4 months).
Students with TP in Grade 3 and performance >50th percentile in Grade 4 formed the
control group (31 students, 20 boys, 11 girls, mean age 9.79  0.3 years). In view of the
highly negatively skewed distribution of GMST-13 scores in Grade 4, scores < 50th
percentile (corresponding to 11/13 correct items) were considered as indicating PD with
simple calculations. In this manner, students in the PD group were defined as those with
very low scores in Grade 3 who maintained below-average performance through Grade 4.
Students in the PD and IP groups were invited to participate in the 8-week intervention
programme that was implemented in either the Spring of Grade 5 or the Fall of Grade 6,
supplementing their regular classroom math lessons. Following additional pre-interven-
tion assessments at Time 3, 13 students in the PD group (8 boys, 5 girls) and 10 students in
the IP group (5 boys, 5 girls) completed the programme and provided complete post-
intervention assessment data at Time 4. Students in the TP group did not attend the
intervention but had baseline (Time 3) and follow-up assessments (at Time 4, i.e., Time 3
plus 8–9 weeks): 26 students (18 boys, 8 girls) in the TP group completed these
assessments. From the 49 students comprising the final sample (13 PD, 10 IP, 26 TP), 18
students (12 boys, 6 girls, mean age 10.79  0.2) were assessed in Grade 5 and 31
students (19 boys, 12 girls, mean age 11.48  0.3) were assessed in Grade 6. A flow chart
of the sampling, assessment, and intervention timeline is shown in Figure 1.
Demographic characteristics of the final sample of participants are described in
Table 1. The three math ability subgroups did not differ on age, gender distribution,
paternal occupation type, or educational level. There was a non-significant trend towards
a greater representation of children from immigrant families in the IP group, although all
students attended Greek school (immersion classrooms) from Grade 1. Maternal
education was, however, higher among students in the TP group (13.67  3.26 years)
as compared to students in the PD group (9.75  4.31 years). Although numbers were
small to permit formal statistical comparisons, there was a tendency for higher
representation of unemployed mothers in the PD group. These findings are not
unexpected in view of the well-established association between maternal education level
and student academic achievement.

Measures
At Time 1 (Grade 3), students were asked to complete the short form of GMST (GMST-13)
consisting 13 simple paper-and-pencil calculations (sample items are presented in
Table 2), involving symbolic number processing, that is, ability to perform mental
computations without using manipulatives, to write down digits, to align numbers
6 Elen S. Karakonstantaki et al.

Time N Tests

Time 1 1382 GMST-13

(Grade 3)

122 122
Low performance Typical performance
(<16th percenle) (>35th percenle)

Time 2 36 31 NUCALC
(Grade 4) WISC>80
TORP >25th
percenle
CBCL< 70

19 17 31 GMST-13
Persistent Improved Typical
Difficulty (PD) Performance (IP) Performance (TP)
(<50th percenle) (>50th percenle) (>50th percenle)

Time 3 19 17 31 GMST-34
(Grade 5 or 6) NUCALC-operaons

INTERVENTION START

Time 4 13 10 26 GMST-34
(Grade 5 or 6) NUCALC-operaons

INTERVENTION END
Figure 1. Flow chart of assessments and timeline of the intervention by participant group.

according to their place value, and to use arithmetic symbols properly. It was developed
and standardized to assess basic calculation abilities according to grade-appropriate Greek
national curriculum guidelines for Grades 1–6 (Papaioannou, Mouzaki, Sideridis, & Simos,
2011).
 Evaluation and remediation of calculation skills 7

Table 1. Demographic characteristics of students in each math ability group

PD (n = 19) IP (n = 17) TP (n = 31)

Mean SD Mean SD Mean SD p-Valuea

Age at T3 (years) 11.14 0.49 11.06 0.70 11.29 0.40 .450

Maternal education (years) 9.75 4.31 13.00 1.53 13.67 3.26 .048*
Paternal education (years) 11.15 4.39 11.29 3.09 13.25 4.39 .520

n % n % N % p-Valuea

Sex
Male 10 52.6 9 52.9 20 64.5 .624
Female 9 47.4 8 47.1 11 35.5
Ethnicity
Greek 19 100.0 14 82.4 30 96.8 .071
Other 0 0.0 3 17.6 1 3.2
Maternal occupation
Clerical with ≥14 years of edu. 3 21.4 5 62.5 3 23.1
Clerical with <14 years of edu. 2 14.3 1 12.5 7 53.8
Unemployed 9 64.3 2 25.0 3 23.1
Paternal occupation
Clerical with ≥14 years of edu. 8 57.1 6 75.0 8 61.5
Clerical with <14 years of edu. 5 35.7 1 12.5 5 38.5
Unemployed 1 7.1 1 12.5 0 0.0

Note. PD = students with persistent math difficulties; IP = students who showed improvement in math
scores performance between Time 1 and Time 2; SD = standard deviation; edu. = education.
Numbers may not correspond to the total number of participants due to missing values.
a
p-Values for Kruskal–Wallis, chi-square, or Fisher’s exact test with Monte Carlo correction.
*p < .05.

At Time 2 (Grade 4), cognitive assessment of students who met the screening criteria
was conducted to exclude intellectual, reading, and behavioural problems. The following
tests were administered: Vocabulary, Block Design, and Digit Span subscales from the
Wechsler Intelligence Scale (WISC-III; Georgas, Paraskevopoulos, Bezevengis, & Gian-
nitsas, 1997; Wechsler, 1991), Word and Pseudoword reading subscales from the Test of
Reading Proficiency (TORP; Padeliadu & Sideridis, 2000), and the Child Behavior
Checklist-Parent Version (Achenbach & Rescorla, 2001). The GMST-13 was re-
administered to assess change in performance over time. To assess the underlying basis
of math performance, the Greek standardized version of Neuropsychological test Battery
for Number Processing and Calculation was used (NUCALC; Koumoula et al., 2004; von
Aster, 2001). As described in more detail in Table 2, NUCALC assesses a wide range of
basic number skills, both non-symbolic (e.g., counting dots) and symbolic (backwards
counting, number comparison, visual absolute and relative quantity estimation, and
problem-solving), and emphasizes translation from one form of number representation to
another (asking students to write down an orally presented numberword in Arabic form).
At Time 3 (Grade 5 or 6), all students were given the full, 34-item version of GMST
(GMST-34) which, in addition to the items included in GMST-13, comprises items
assessing higher level operations (multidigit calculations and calculations involving
fractions and decimals), without time constraint. While GMST-13 was developed as a brief
8 Elen S. Karakonstantaki et al.

Table 2. Description of math assessments

Test Description
NUCALC Students. . .

Dot counting . . .count arrays of dots and say or write the number
Counting backwards . . .count backwards from 22 to 1 and from 67 to 54
Number dictation . . .write an orally presented numeral (14, 1,200, 503, 4,658)
Mental/written addition . . .mentally compute the sum of orally or visually presented
numbers (5 + 3, 15 + 12, 14 + 8, 17 + 25)
Mental/written subtraction . . .mentally compute the difference of orally or visually presented
numbers (17 5, 14 6, 32 17)
Mental multiplication . . .mentally compute the product of orally presented numbers
(3 9 2, 2 9 6, 4 9 4)
Reading numbers . . .name printed numbers (15, 1,900, 305)
Number line (oral) . . .indicate the position of an orally presented number on a 0–100
number line
Number line (written) . . .indicate the position of a printed number on a 0–100 number line
Oral number comparison . . .determine the largest of two orally presented numbers
(49 or 51, 2,009 or 2,090)
Written number comparison . . .determine the largest of two printed numbers (49 or 51, 2,009
or 2,090)
Perceptive estimation . . .estimate the quantity of object arrays (57 balls, 89 cups) each
displayed for 5 s
Contextual estimation . . .rate described quantities as large, medium, or small (8 lamps in a
room, 2 children in a family, 2 clouds in the sky)
Problem-solving . . .mentally compute orally presented problems (Ann has got 5 marbles.
Peter has got 3 less than Ann. How many marbles does Peter have?)
GMST
13-item form . . .solve 13 simple math problems in paper and pencil (1 + 1, 5
1, 32 + 24 + 40, 36 15, 68 + 23, 33 17, 16/4)
34-item form . . .solve 34 simple and advanced math problems in paper and
pencil (1 + 1, 36 15, 229 + 5,048 + 63 + 1,381, 15 5 = . . ., 3 + 3,
1 1

823 9 96, 614 + 158 + 412, 2 . . . = 14)

Note. NUCALC = Neuropsychological test Battery for Number Processing and Calculation;
GMST = Greek Math Screening Test.
Examples are given in parentheses.

screening tool for classroom administration, GMST-34 was designed as an achievement

test and is normally distributed in the population. By design, GMST-13 is characterized by
very high sensitivity (100%) and moderate specificity (75%) (compared against a cut-
off = 30/31 points on GMST-34, corresponding to the 25th percentile of the normative
distribution). In addition, the timed addition, subtraction, and multiplication tasks from
NUCALC were administered.
At Time 4 (Grade 5 or 6; following intervention for PD and IP groups), math assessment
was repeated using the same tests as at Time 3.

Intervention procedure
MyMath intervention programme consisted of 16 lessons on basic number knowledge,
usually mastered by Grade 3. Lessons lasted 1" hr each and were performed in small
 Evaluation and remediation of calculation skills 9

groups twice a week for 8 weeks, by a child psychologist (first author) with experience in
the diagnosis and remediation of learning disabilities. The same person organized the
intervention and conducted the scripted lessons (a sample lesson is provided in Table 3).
Intervention focused on counting and decomposition strategies, place value, and
addition/subtraction problems involving currency and time (see Appendix for a detailed
description of the intervention procedure for each domain of numeracy). Emphasis was
given on quantity representation promoting number sense and conceptual understanding
of operations as proposed in the literature (Butterworth, 2010; Kaufmann et al., 2003).
For instance, 3 + 5 = 8 is a symbolic representation of an addition fact, but students have
to be able to understand that it refers to grouping three objects and five objects together,
before being able to automatically recall the answer. The instructional approach was

Table 3. Lesson plan for the concept of place value

Exercise Task Examples

1 Using specialized material (coloured number rods 25, 78, 138, 215, 1,471, 1,354, 200,
representing ones, tens, hundreds, and thousands), 500, 80, 40, 1,056, 1,038, 1,304,
students are asked to write down the Arabic number 1,501
represented by the number rods
2 Students are asked to represent orally presented 5, 9, 32, 88, 154, 1,261, 111, 70,
numbers placing number rods in the appropriate 105, 430, 1,026, 1,480, 1,307, 1,408
columns on a board with discrete columns for ones,
tens, hundreds, and thousands
3 Students are asked to represent orally presented 138, 256, 1,034, 1,062 and other
numbers using number rods without the aid of the numbers proposed by students
place value board. The goal is to grasp the idea that
any place within a multidigit number (including 0)
represents a value
4 Students are asked to tell how many ones, tens, 5, 50, 500, 5,000
hundreds, or thousands numbers consist 38, 380, 308, 3,008
5 Students are asked to decide whether certain digits 25, 50, 500, 5,271
represent ones, tens, hundreds, or thousands 86, 871, 380, 1,458
6 Students are asked to write in arabic form the 5 Th, 3 T, 4 O, 1 Th, 5 O
numbers representing given thousands (Th),
hundreds (H), tens (T), and ones (O)
7 Students are asked to write in Arabic form the 1,000 + 300 + 50 + 6
multidigit number that is composed of other given 700 + 9
numbers
8 Students are asked to break down numbers into their 1,256 as: 1,000 + 200 + 50 + 6
components or 1 Th, 2 H, 5 T, and 6 O
9 Students are asked to represent written One thousand and sixty-two and
numberwords using number rods others proposed by students
10 Students are presented with four digits and are asked 0, 3, 5, 8: 8,530 and 358
to arrange them so as to form the largest and the
lowest possible numerical values
11 Stamps of ones, tens, hundreds, and thousands are 1,256, 4,538, 2,741, 2,346, 2,205,
provided representing numbers as quantities, and 1,301, 300, 14,000
students are asked to write them in Arabic form
12 Students are asked to use the stamps to represent
orally presented numbers as quantities
10 Elen S. Karakonstantaki et al.

explicit and implemented commercially available products for hands-on activities (e.g.,
manipulatives for place value). Lessons further involved conversations on everyday life
situations requiring mathematical thinking, encouraging teamwork, and providing
students ample time to reflect on their answers. Direct teaching was used in some cases,
as in the lesson on problem-solving which begun by presenting students with a number of
steps to compute a solution. Still students had the opportunity to work in groups to solve
problems, were asked to construct their own problems, and were encouraged to actively
engage in problem-solving tasks rather than simply reproducing the procedure modelled
for them by the instructor.
Throughout the lessons, it was made regular use of interactive math games (using dice,
colouring photocopies, addition and subtraction cards, puzzles requiring mathematic
thought) to enhance sense of self-efficacy and motivation to engage in math activities
(Ormrod, 2003) for these students who typically had a long history of frustrating
experiences associated with math coursework.

Statistical analysis
In the present analyses, the outcome of interest consisted of students’ scores on
neuropsychological and psychoeducational tests. Data analysis was performed using
the statistical software SPSS version 21 (SPSS Inc., Chicago, IL, USA). Bivariate
associations between categorical variables were assessed using Pearson’s chi-square
test or Fisher’s exact test with Monte Carlo correction. As the neuropsychological and
psychoeducational scores failed the normality test (Kolmogorov–Smirnov test with
Lilliefors significance correction), nonparametric tests were applied to assess the
association between the neuropsychological and psychoeducational scores categorical
(Kruskal–Wallis and Mann–Whitney U-tests) variables. Also, the Wilcoxon signed-rank
test was used to assess differences in pre- and post-intervention performance. The
significance level was set at .05 (two-sided). Effect sizes using Cohen’s d were also
calculated.

Results
Assessments at Time 2
Table 4 indicates that at Time 2, on average, students in the PD group scored lower on
WISC-III subtests than the IP group. These differences were associated with medium/large
effect sizes reaching significance for Vocabulary (p = .044). Interestingly, these effects
appeared to be largely due to lower maternal education of the PD group, and therefore
may be attributed to reduced cultural opportunities faced by children in this group.
Compared to the typically achieving group, students in the PD group showed significantly
lower performance on all WISC subscales (p < .05; data not shown). Still, individual
scores on Block design and Vocabulary remained in the average range (<1 SD below
mean) indicating typical general cognitive ability. The IP and TP groups showed
comparable performance on WISC subscales with the exception of Digit Span (IP < TP,
p = .035, data not shown) and, again, individual scores did not fall lower than 1 SD below
the population mean. It should be noted, however, that group differences on Digit Span
scores may indeed reflect weaker verbal short-term and working memory but may also
reflect deficient symbolic number processing. No differences were found in measures of
reading proficiency (TORP) between students in the PD and IP groups. Both math-
impaired groups had comparable performance to the TP group on word reading accuracy
 Evaluation and remediation of calculation skills 11

Table 4. Neuropsychological and psychoeducational test scores for PD and IP groups at Time 2

PD (n = 19) IP (n = 17)

Measure Mean SD Mean SD p-Valuea Cohen’s d

WISC-III
Vocabulary 8.58 3.22 10.94 2.25 .044* 0.85
Block design 7.95 2.46 9.65 2.29 .087 0.72
Digit span 6.11 2.58 7.94 2.19 .139 0.76
TORP
Pseudowords 24.11 8.27 25.06 7.18 .925 0.12
Words 54.89 16.38 61.41 15.48 .558 0.41
NUCALC (accuracy)
Dot counting 1.58 0.69 1.59 0.62 .999 0.02
Counting backwards 2.74 1.37 3.47 0.87 .085 0.64
Number dictation 1 14.26 2.96 14.00 2.72 .936 0.09
Mental addition 11.16 3.96 14.71 9.46 .157 0.49
Written addition 12.21 3.33 15.38 0.96 <.001* 1.29
Mental subtraction 8.26 3.75 11.94 8.55 .107 0.56
Written subtraction 8.00 3.46 12.25 4.00 .001* 1.14
Mental multiplication 8.79 3.34 11.35 3.48 .017* 0.75
Reading numbers 14.21 2.42 15.29 1.21 .088 0.56
Number line (oral) 4.42 1.71 5.29 0.99 .084 0.62
Number line (written) 4.42 1.95 5.53 0.87 .020* 0.74
Oral number comparison 13.05 2.07 13.94 1.64 .298 0.48
Written number comparison 17.16 2.93 18.71 1.72 .075 0.65
Perceptive estimation 4.32 1.67 4.71 1.57 .784 0.24
Contextual estimation 14.00 4.37 18.00 2.12 <.001* 1.16
Problem-solving 6.05 4.20 9.35 2.64 .004* 0.94
NUCALC speed (s)
Dot counting 18.89 5.53 15.47 4.42 .077 0.68
Counting backwards 34.00 13.49 28.18 10.76 .304 0.48
Number dictation 43.89 18.39 37.76 10.23 .306 0.41
Mental addition 98.32 50.98 61.18 38.92 .017* 0.82
Mental subtraction 170.84 67.52 111.59 50.76 .015* 0.99
Mental multiplication 33.86 21.05 15.94 12.09 .001* 1.03
Reading numbers 23.94 11.28 17.47 3.59 .063 0.77
Number line (oral) 8.47 5.31 9.00 9.14 .967 0.07
Number line (written) 7.79 3.98 6.63 2.19 .725 0.36
Oral number comparison 44.95 8.02 43.81 6.16 .928 0.16
Written comparison 35.68 11.79 29.00 7.93 .116 0.66
Contextual estimation 48.79 13.29 47.38 11.25 .938 0.11
Problem-solving 88.58 57.38 66.47 37.92 .323 0.45

Note. PD = students with persistent math difficulties; IP = students who showed improvement in math
scores performance between Time 1 and Time 2; WISC-III = Wechsler Intelligence Scale, 3rd ed.;
TORP = Test of Reading Proficiency; NUCALC = Neuropsychological test Battery for Number
Processing and Calculation; Cohen’s d values >0 indicate higher performance (increased accuracy or
speed) for the IP group.
Effect sizes indicating significant differences are presented in boldface.
a
p-Values for Mann–Whitney U-test for two independent samples.
*p < .05.
12 Elen S. Karakonstantaki et al.

(PD vs. TP: p = .112; IP vs. TP: p = .690; data not shown), but scored lower on
pseudoword reading accuracy (PD vs. TP: p = .001; IP vs. TP: p = .005; data not shown).
As expected, students in the PD group showed reduced accuracy than typically
achieving students on all NUCALC subscales. The difference was significant in every
subscale (p < .005; 0.15 < d < 2.25; data not shown) except for dot counting (p = .561),
number dictation (p = .098), mental addition (p = .124) (although significant, more time
was needed to perform these tasks), and perceptive estimation (p = .863). Completion
times were systematically higher on all subscales, except for perceptive estimation where
a time limit of 5 s was used.
Compared to the IP group, students with PD showed significantly lower accuracy
on the following tasks, with large associated effect sizes: written addition (Mann–
Whitney U = 68.00, p = .001, d = 1.29), written subtraction (U = 65.50, p = .001,
d = 1.14), contextual estimation (U = 76.00, p = .001, d = 1.16), and problem-solving
(U = 86.50, p = .004, d = 0.94; see Table 4). Significant, medium-effect-size differ-
ences in the same direction were noted on the written number line (U = 113.00,
p = .020, d = 0.74) and mental multiplication tasks (U = 83.00, p = .017, d = 0.75).
Smaller, yet medium-size differences on counting backwards (U = 112.00, p = .085,
d = 0.64), oral number line (U = 118.50, p = .084, d = 0.62), and written number
comparison (U = 112.50, p = .075, d = 0.65) were also noted, not reaching signifi-
cance. Task completion times, listed in Table 4, do not suggest the presence of a
speed–accuracy trade-off in the performance of students in the PD group, who
generally required longer time to complete the NUCALC tasks. The difference between
the two groups on speed reached significance on the calculation tasks (addition,
U = 95.00, p = .017, d = 0.82; subtraction, U = 83.00, p = .015, d = 0.99; and
multiplication, U = 68.00, p = .001, d = 1.03).

Assessments at Time 3
On the third assessment, immediately prior to enrolling in the study intervention (Time 3),
students in the IP group scored higher than students in the PD group on GMST-34. This
difference only approached significance (U = 44.00, p = .05) despite being associated
with a large effect size (d = .96).

Intervention effects
Compared to their pre-intervention scores (Time 3), improvement was found among
students in the PD group on virtually every measure, with pre- versus post-intervention
differences reaching significance for GMST-34 (Wilcoxon test; z = 2.47, p = .014,
d = 0.37) and multiplication speed (z = 2.60, p = .009, d = 0.54; see Table 5). These
differences were associated with small-to-medium effect sizes.
On average, the pattern of changes in performance between Time 3 and Time 4 of
students in the IP groups was variable, ranging from small-size improvement (e.g., on
subtraction speed; d = 0.32) to small-size reduction (e.g., on addition and multiplication
accuracy; d = 0.43 and d = 0.40, respectively). Task performance remained essentially
unchanged among students in the TP group, with a small, yet significant improvement on
GMST-34 (p = .049, d = 0.31), during this 8-week period.
Following intervention (Time 4), the difference between PD and IP groups on GMST-
34 was smaller in size (d = 0.68 vs. d = 0.96 at Time 3), failing to reach significance
(p = .10).
 Evaluation and remediation of calculation skills 13

Discussion
Math skill profiles at baseline
As shown in Table 4, students in the PD group underperformed students who did not
experience persistent math difficulties primarily on tasks assessing symbolic processes
(written addition/subtraction, mental multiplication, problem-solving, and placing
visually presented values [Arabic digits] on a number line). Among them, paper-and-
pencil exercises require efficient processing of number symbols according to their place
value. Mental multiplication is characterized by a greater degree of abstractness than
addition and subtraction and involves a concept that does not derive as intuitively
from concrete life experiences. Thus, it may prove especially challenging for students
who experience difficulties in symbolic processes as well as poor retrieval capacity of
simple math facts. Moreover, mental addition and subtraction appeared to place
increased processing demands on students with persistent math difficulties, as

Table 5. Performance accuracy and speed (s) on pre (Time 3)- and post-intervention (Time 4)
assessments

Time 3 Time 4

Group Test Mean SD Mean SD p-Valuea Cohen’s d

PD (n = 13) GMST-34 score 13.81 4.20 15.50 4.80 .014* 0.37

Addition score 11.63 3.50 12.43 2.59 .165 0.39
Addition speed 72.69 32.08 58.79 27.27 .054 0.47
Subtraction score 8.13 3.67 9.86 3.37 .180 0.49
Subtraction speed 122.94 53.56 108.21 40.05 .362 0.31
Multiplication score 10.00 2.50 11.00 2.08 .141 0.43
Multiplication speed 28.44 22.54 19.21 8.63 .009* 0.54
IP (n = 10) GMST-34 score 17.73 3.98 18.80 4.87 .190 0.24
Addition score 13.73 2.20 12.60 2.95 .590 0.43
Addition speed 45.18 29.17 46.50 20.46 .953 0.05
Subtraction score 14.55 11.40 10.60 3.31 .541 0.47
Subtraction speed 101.45 69.09 83.20 40.24 .066 0.32
Multiplication score 14.64 10.43 11.70 0.67 .854 0.40
Multiplication speed 15.36 12.45 16.30 12.45 .570 0.08
TP (n = 26) GMST-34 score 24.81 3.83 25.96 3.62 .049* 0.31
Addition score 14.59 1.65 14.85 1.52 .692 0.16
Addition speed 32.93 22.60 34.35 33.04 .736 0.05
Subtraction score 13.07 2.42 13.65 2.04 .331 0.26
Subtraction speed 56.59 36.81 51.65 43.13 .089 0.12
Multiplication score 11.96 0.19 12.15 0.78 .180 0.33
Multiplication speed 8.67 2.47 8.04 2.24 .194 0.27

Note. Time 3 (T3) = pre-intervention assessment; Time 4 (T4) = post-intervention assessment;

PD = students with persistent math difficulties; IP = students who showed improvement in math
scores performance between Time 1 and Time 2; TP = typically achieving students; GMST = Greek Math
Screening Test.
Effect sizes indicating significant differences are presented in boldface. Cohen’s d values >0 indicate
improved T4 versus T3 performance (increased accuracy or speed).
a
p-Values Wilcoxon signed-rank test.
*p < .05.
14 Elen S. Karakonstantaki et al.

indicated by significantly prolonged completion times. The latter finding may be

accounted for by difficulties in recalling simple math facts, as well as by somewhat
reduced short-term and working memory capacity (associated with medium effect sizes).
Solving arithmetic problems is a complex abstract reasoning task, requiring access to the
quantity representation of numberwords and requiring intact language skills as well.
Conversely, students with PD performed comparable to IP students when asked to place
an orally presented value (numberword) on a number line, as they can probably rely on
linguistic symbols. However, when the value is visually presented, (Arabic) digits must
first be transformed to their corresponding quantities, posing increased difficulties to PD
students. In a longitudinal study, Geary et al. (2012) also observed a quantity
representation deficit among students with math difficulties.
Although not conclusive, the present findings are consistent with a dissociation
between symbolic and non-symbolic quantity representation and processing (Wilson
& Dehaene, 2007) that best describes the cognitive profile of students with
persistent difficulties in the present study. A number sense deficit is not
corroborated by our results, as students with PD managed to handle exact
numerosity, adequately performed number comparison, and approximate estimation.
Other researchers have come to a similar conclusion assessing performance on
comparison tasks (Mussolin et al., 2010; Rousselle & Noёl, 2007). The profile of the
PD group resembles that of students with poor numeracy skills, who show a deficit
in relating symbols to quantities and fail on most symbolic tasks although they may
sometimes manage to compensate for their difficulties in certain simpler symbolic
tasks, such as naming Arabic numerals (Chan et al., 2013). Interestingly, the
purported deficit in symbolic processing skills was not observed on the TORP Word
Reading task, as no differences in performance were noticed between the three
groups. The tenuous association between low numeracy and performance in
decoding tasks is also consistent with Dowker (2016). However, both PD and IP
groups scored significantly lower than students in the TP group in pseudoword
decoding. A possible explanation could be that students in the PD and IP groups
compensated for their symbolic processing deficits by relying on phonological
representations when possible (e.g., for numberwords and words alike; De Smedt,
Taylor, Archibald, & Ansari, 2010). Szucs et al. (2013) failing to detect deficits in
number processing skills among students with dyscalculia implicated deficits in
visuospatial short-term and working memory, as well as in inhibition capacity,
instead of a difficulty in accessing non-symbolic information (quantity) from symbolic
information (Arabic digits). Unfortunately, these cognitive abilities were not assessed
in the present study.
The IP group did not differ significantly from the TP group on NUCALC scores, failing
to reveal an underlying deficit in specific math domains. The lower scores of the IP group
on the extended version of the math achievement test at Times 3 and 4 (GMST-34)
compared to typically achieving students may be due to poorer verbal short-term/working
memory (Digit Span scores). A short-term/working memory deficit has often been
proposed to be responsible for calculation difficulties (Geary, 1993; Wilson & Dehaene,
2007; Wilson & Swanson, 2001).

Effects associated with math intervention

Contrary to our expectations, intervention effects were more prominent for the PD as
compared to the IP group. This finding is in accordance with the notion that children with
 Evaluation and remediation of calculation skills 15

diverse profiles respond differently to remedial programmes (Dowker & Sigley, 2010).
Students in the former group showed increased post-intervention scores and shorter
reaction times on virtually every measure, associated with small-to-medium effect sizes
(see Table 5). GMST-34 is a formal test assessing complex computational skills, where
performance of students with severe difficulties unlike those with TP is probable to
deteriorate with time. The statistically significant improvement on GMST-34 scores is an
important finding, indicative of the possible efficiency of interventions focusing on
conceptual teaching fostering basic arithmetic understanding for older students facing
calculation difficulties, as suggested by previous research (Scherer et al., 2016). We argue
that our results imply a transfer of knowledge derived from practice in non-symbolic tasks
(quantity manipulation) to symbolic ones (written operations) for students with
persistent math difficulties as previously suggested for young students with typical math
achievement (Clements & Sarama, 2011; Hyde et al., 2014; Rousselle & Noёl, 2007).
Conceptual teaching focuses on building concepts gradually, beginning with concrete
and moving to more abstract materials, following the physical process of development of
symbolic thought. Evidence exists that this method may benefit students with a long
history of math difficulties even in more complex mathematic tasks (Witzel, 2005).
Participants in the present intervention programme, despite their age, were given the
opportunity to use manipulatives to represent quantity and basic operations, in order to
promote number sense, and facilitate grasping of the symbolic nature of numbers. They
were also encouraged to make inferences about everyday life situations requiring addition
or subtraction and test their assumptions, a skill that may be distinct from the ability to use
algorithms (Carraher et al., 1985). This focus on conceptual understanding, combined
with some procedural training (e.g., in calculation strategies and problem-solving), has
benefited students with persistent difficulties in our sample, as also suggested by other
researchers (Scherer et al., 2016).
The magnitude of test score improvement was encouraging, especially when taking
into account (1) the persistent nature of students’ calculation difficulties, (2) the short
duration of the programme, (3) the relatively advanced grade when intervention was
implemented, and (4) the group-based delivery of the intervention. A more intense/
systematic intervention delivered over a longer period may have led to more robust
results. Including an active control group of students with persistent math difficulties who
would receive a conceptually different type of intervention is required to establish the
effectiveness of the current programme. Furthermore, a more thorough neuropsycho-
logical assessment, including tests of executive function and visuospatial processing, may
have revealed additional deficits among students experiencing math difficulties (Cowan &
Powell, 2013; Karagiannakis, Baccaglini-Frank, & Papadatos, 2014; Szucs et al., 2013). A
memory measure that does not involve numeric stimuli would have been more
appropriate for students with math learning difficulties.
We chose to define our experimental groups in terms of persistence of impaired
performance in arithmetic tests instead of traditional definitions of math disability, that is,
poor math achievement alone or IQ-math achievement discrepancy. The characteristics
of our PD group are consistent with the suggested definition of MD by Mazzocco and
Myers (2003) requiring severe math underperformance persisting for two or more years.
The degree of persistence of math difficulties in our study resembles that of Shalev, Manor,
Auerbach, and Gross-Tsur (1998) given that approximately 50% of low-achieving students
at Time 1 in our study (n = 36) showed persistent difficulties at Time 2, while the
remaining demonstrated IP (although as a group they still underperformed students in the
TP group). Although a discrepancy criterion was not used in the present study, the PD
16 Elen S. Karakonstantaki et al.

group scored significantly lower than the IP and TP groups on WISC-III Vocabulary, Block
Design, and Digit Span scales. The (smaller) difference between the PD and IP groups
could be largely attributed to family/cultural factors, associated with maternal education.
However, the much more pronounced underperformance of the PD group as compared
to typically math-achieving students is consistent with the view that students with low
math achievement tend to score lower on IQ measures compared to both students
meeting the IQ-achievement discrepancy criterion for math learning disability and
students with typical achievement. Moreover, students with poor math performance are
found to perform poorly on general cognitive tasks due to the strong correlations between
math performance and general cognitive abilities (Tolar, Fuchs, Fletcher, Fuchs, &
Hamlett, 2016). Other researchers have also reported significantly lower IQ scores for
students with math learning disabilities compared to students with typical achievement
(Geary et al., 2012).
The present study suggests that students, who struggle with basic calculation at the
end of elementary school, may benefit in certain areas of math reasoning from conceptual
interventions that provide explicit teaching of basic concepts, as well as opportunities to
interact with the physical environment and to construct knowledge through experience,
while providing basic procedural training. Our results further suggest that persistent
calculation difficulties may arise from underlying deficits in relating symbolic and non-
symbolic number representation and processing. We argue that conceptual interventions
aiming to enhance quantity representation contribute to the organization of a semantic
network where numeric associations become meaningful.

References
Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the ASEBA school-age forms & profiles.
Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families.
Butterworth, B. (2005). The development of arithmetical abilities. Journal of Child Psychology and
Psychiatry, 46, 3–18. https://doi.org/10.1111/j.1469-7610.2005.00374.x
Butterworth, B. (2010). Foundational numerical capacities and the origins of dyscalculia. Trends in
Cognitive Sciences, 14, 534–541. https://doi.org/10.1016/j.tics.2010.09.007
Butterworth, B., Varma, S., & Laurillard, D. (2011). Dyscalculia: From brain to education. Science,
332, 1049–1053. https://doi.org/10.1126/science.1201536
Carraher, T. N., Carraher, D. W., & Schliemann, A. D. (1985). Mathematics in the streets and in
schools. British Journal of Developmental Psychology, 3, 21–29. https://doi.org/10.1111/j.
2044-835X.1985.tb00951.x
Chan, W. W. L., Au, T. K., & Tang, J. (2013). Developmental dyscalculia and low numeracy in
Chinese children. Research in Developmental Disabilities, 34, 1613–1622. https://doi.org/10.
1016/j.ridd.2013.01.030
Chodura, S., Kuhn, J.-T., & Holling, H. (2015). Interventions for children with mathematical
€ Psychologie, 223, 129–144. https://doi.org/10.1027/
difficulties. A meta-analysis. Zeitschrift fur
2151-2604/a000211
Clements, D. H., & Sarama, J. (2011). Early childhood mathematics intervention. Science, 333, 968–
970. https://doi.org/10.1126/science.1204537
Cowan, R., & Powell, D. (2013). The contributions of domain-general and numerical factors to third
grade arithmetic skills and mathematical learning disability. Journal of Educational Psychology,
106, 214–229. https://doi.org/10.1037/a0034097
De Smedt, B., & Gilmore, C. K. (2011). Defective number module or impaired access? Numerical
magnitude processing in first graders with mathematical difficulties. Journal of Experimental
Child Psychology, 108, 278–292. https://doi.org/10.1016/j.jecp.2010.09.003
 Evaluation and remediation of calculation skills 17

De Smedt, B., Taylor, J., Archibald, L., & Ansari, D. (2010). How is phonological processing related to
individual differences in children’s arithmetic skills? Developmental Science, 13, 508–520.
https://doi.org/10.1111/j.1467-7687.2009.00897.x
Dehaene, S., Piazza, M., Pinel, P., & Cohen, L. (2003). Three parietal circuits for number processing.
Cognitive Neuropsychology, 20, 487–506. https://doi.org/10.1080/02643290244000239
Dowker, A. (2001). Numeracy recovery: A pilot scheme for early intervention with young children
with numeracy difficulties. Support for Learning, 16, 6–10. https://doi.org/10.1111/1467-
9604.00178
Dowker, A. (2016). Factors that influence improvement in numeracy, reading and comprehension
in the context of a numeracy intervention. Frontiers in Psychology, 7, 1–10. https://doi.org/10.
3389/fpsyg.2016.01929
Dowker, A., & Sigley, G. (2010). Targeted interventions for children with arithmetical difficulties.
British Journal of Educational Psychology, Monograph Series II, 7, 65–81. https://doi.org/10.
1348/97818543370009x12583699332492
Engelmann, S., Carnine, D., Kelly, B., & Engelmann, O. (1991–1995). Connecting math concepts.
Chicago, IL: Science Research.
Fuchs, L. S., Powell, S. R., Seethaler, P. M., Cirino, P. T., Fletcher, J. M., Fuchs, D., . . . Zumeta, R. O.
(2009). Remediating number combination and word problem deficits among students with
mathematics difficulties: A randomized control trial. Journal of Educational Psychology, 101,
561–576. https://doi.org/10.1037/a0014701
Geary, D. C. (1993). Mathematical disabilities: Cognitive, neuropsychological and genetic
components. Psychological Bulletin, 114, 345–362. https://doi.org/10.1037/0033-2909.114.
2.345
Geary, D. C., & Hoard, M. K. (2005). Learning disabilities in arithmetic and mathematics. Theoretical
and empirical perspectives. In J. I. D. Campbell (Ed.), Handbook of mathematical cognition
(pp. 253–267). New York, NY and Hove, UK: Psychology Press.
Geary, D. C., Hoard, M. K., Nugent, L., & Bailey, D. H. (2012). Mathematical cognition deficits in
children with learning disabilities and persistent low achievement: A five-year prospective study.
Journal of Educational Psychology, 104, 206–223. https://doi.org/10.1037/a0025398
Georgas, D. D., Paraskevopoulos, Ι. Ν., Bezevengis, E. G., & Giannitsas, Ν. D. (1997). Εlleniko WISC-
IIΙ: Wechsler klimakes noymosenis gia paidia. [Greek WISC-III: Wechsler Intelligence Scales
for children]. Athens, Greece: Εllenika Grammata.
Great Source. (2006). Every Day Counts grades K-6: Research base and program effectiveness.
(Computer manual). Retrieved from http://www.greatsource.com/GreatSource/pdf/EveryDa
yCountsResearch206.pdf
Griffin, S. (2004). Building number sense with Number Worlds: A mathematics program for young
children. Early Childhood Research Quarterly, 19, 173–180. https://doi.org/10.1016/j.ecresq.
2004.01.012
Holmes, W., & Dowker, A. (2013). Catch Up Numeracy: A targeted intervention for children who are
low-attaining in mathematics. Research in Mathematics Education, 15, 249–265. https://doi.
org/10.1080/14794802.2013.803779
Hyde, C. D., Khanum, S., & Spelke, E. S. (2014). Brief non-symbolic, approximate number practice
enhances subsequent exact symbolic arithmetic in children. Cognition, 131, 92–107. https://
doi.org/10.1016/j.cognition.2013.12.007
Jones, E. D., Wilson, R., & Bhojwani, S. (1998). Mathematics instruction for secondary students with
learning disabilities. In D. P. Rivera (Ed.), Mathematics education for students with learning
disabilities (pp. 155–176). Austin, TX: Pro-Ed.
Karagiannakis, G., Baccaglini-Frank, A., & Papadatos, Y. (2014). Mathematical learning difficulties
subtypes classification. Frontiers in Human Neuroscience, 8, 1–5. https://doi.org/10.3389/
fnhum.2014.00057
Kaufmann, L., Handl, P., & Th€ ony, B. (2003). Evaluation of a numeracy intervention program
focusing on basic numerical knowledge and conceptual knowledge. A pilot study. Journal of
Learning Disability, 36, 564–573. https://doi.org/10.1177/00222194030360060701
18 Elen S. Karakonstantaki et al.

Kaufmann, L., Mazzocco, M. M., Dowker, A., von Aster, M., Gӧbel, S. M., Grabner, R. H., . . . Nuerk, H.-
C. (2013). Dyscalculia from a developmental and differential perspective. Frontiers in
Psychology, 4, 1–5. https://doi.org/10.3389/fpsyg.2013.00516
Kaufmann, L., & von Aster, M. (2012). The diagnosis and management of dyscalculia. Deutsches
€
Arzteblatt International, 109, 767–778. https://doi.org/10.3238/arztebl.2012.0767
Ketterlin-Geller, L. R., Chard, D. J., & Fien, H. (2008). Making connections in mathematics.
Conceptual mathematics intervention for low-performing students. Remedial and Special
Education, 29, 33–45. https://doi.org/10.1177/0741932507309711
Koumoula, A., Tsironi, V., Stamouli, V., Bardani, I., Siapati, S., Graham, A., . . . von Aster, M. (2004).
An epidemiological study of number processing and mental calculation in Greek schoolchildren.
Journal of Learning Disabilities, 37, 377–388. https://doi.org/10.1177/00222194040370
050201
Landerl, K., Bevan, A., & Butterworth, B. (2004). Developmental dyscalculia and basic numerical
capacities: A study of 8-9 year-old students. Cognition, 93, 99–125. https://doi.org/10.1016/j.c
ognition.2003.11.004
Lee, M. J., & Tingstron, D. H. (1994). A group math intervention: The modification of Cover, Copy
and Compare for group application. Psychology in the Schools, 31, 133–145. https://doi.org/10.
1002/1520-6807(199404)31:2 < 133:aid-pits2310310208 > 3.0.co;2-g
Mazzocco, M. M. M., & Myers, G. F. (2003). Complexities in identifying and defining Mathematics
Learning Disability in the primary school-age years. Annals of Dyslexia, 53, 218–253. https://
doi.org/10.1007/s11881-003-0011-7
Miller, S. P., & Mercer, C. D. (1998). Educational aspects of mathematics disabilities. In D. P. Rivera (Ed.),
Mathematics education for students with learning disabilities (pp. 81–96). Austin, TX: Pro-Ed.
Montague, M. (1998). Cognitive strategy instruction in mathematics for students with learning
disabilities. In D. P. Rivera (Ed.), Mathematics education for students with learning disabilities
(pp. 177–199). Austin, TX: Pro-Ed.
Mussolin, C., Mejias, S., & Noёl, M.-P. (2010). Symbolic and nonsymbolic number comparison in
children with and without dyscalculia. Cognition, 115, 10–25. https://doi.org/10.1016/
j.cognition.2009.10.006
Ormrod, J. E. (2003). Educational psychology: Developing learners. Upper Saddle River, NJ:
Merrill/Prentice Hall.
Padeliadu, S., & Sideridis, G. (2000). Discriminant validation of the Test of Reading Performance
(TORP) for identifying children at risk of reading difficulties. European Journal of Psychological
Assessment, 16, 139–146. https://doi.org/10.1027//1015-5759.16.2.139
Papaioannou, S., Mouzaki, A., Sideridis, G., & Simos, P. (2011). The screening test of mathematical
achievement for elementary school students. Proceedings of the Second Greek National
Conference of the Special Education Association: Special education as a venue for progress in
science and practice. Athens, Greece: Grigoris Publications. (In Greek).
Rӓsӓnen, P., Salminen, J., Wilson, A. J., Aunio, P., & Dehaene, S. (2009). Computer-assisted
intervention for children with low numeracy skills. Cognitive Development, 24, 450–472.
https://doi.org/10.1016/j.cogdev.2009.09003
Rousselle, L., & Noёl, M.-P. (2007). Basic numerical skills in children with mathematics learning
disabilities: A comparison of symbolic vs non-symbolic number magnitude processing.
Cognition, 102, 361–395. https://doi.org/10.1016/j.cognition.2006.01.005
Scherer, P., Beswick, K., DeBlois, L., Healy, L., & Moser Opitz, E. (2016). Assistance of students with
mathematical learning difficulties: How can research support practice? ZDM, 48, 633–649.
https://doi.org/10.1007/s11858-016-0800-1
Shalev, R. S., Manor, O., Auerbach, J., & Gross-Tsur, V. (1998). Persistence of developmental
dyscalculia: What counts? Results from a 3-year prospective follow-up study. The Journal of
Pediatrics, 133, 358–362. https://doi.org/10.1016/S0022-3476(98)70269-0
Silver, C. H., Pennett, H. D.-L., Black, J. L., Fair, G. W., & Balise, R. R. (1999). Stability of arithmetic
disability subtypes. Journal of Learning Disabilities, 32, 108–119. https://doi.org/10.1177/
002221949903200202
 Evaluation and remediation of calculation skills 19

Szucs, D., Devine, A., Soltesz, F., Nobes, A., & Gabriel, F. (2013). Developmental dyscalculia is
related to visuo-spatial memory and inhibition impairment. Cortex, 49, 2674–2688. https://doi.
org/10.1016/j.cortex.2013.06.007
Tolar, T. D., Fuchs, L., Fletcher, J. M., Fuchs, D., & Hamlett, C. L. (2016). Cognitive profiles of
mathematical problem solving learning disability for different definitions of disability. Journal of
Learning Disabilities, 49, 240–256. https://doi.org/10.1177/0022219414538520
von Aster, M. (2001). Die neuropsychologische testbatterie fur zahlenverarbeitung und rechnen
bei kindern (ZAREKI). [The neuropsychological test battery for number processing and
calculation in children: NUCALC]. Frankfurt, Germany: Swets and Zeitliger.
Wechsler, D. (1991). The Wechsler intelligence scale for children-third edition (WISC-III). San
Antonio, TX: The Psychological Corporation.
Wilson, A. J., & Dehaene, S. (2007). Number sense and developmental dyscalculia. In D. Coch, K. W.
Fisher & G. Dawson (Eds.), Human behavior, learning and the developing brain. Atypical
development (2nd ed., pp. 212–238). New York, NY: Guilford Press.
Wilson, K. M., & Swanson, H. L. (2001). Are mathematics disabilities due to a domain-general or a
domain-specific working memory deficit? Journal of Learning Disabilities, 34, 237–248.
https://doi.org/10.1177/002221940103400304
Witzel, B. S. (2005). Using CRA to teach algebra to students with math difficulties in inclusive
settings. Learning Disabilities: A Contemporary Journal, 3, 49–60.
Wright, R. J. (2003). A mathematics recovery: Program of intervention in early number learning.
Australian Journal of Learning Disabilities, 8, 6–11. https://doi.org/10.1080/1940415030
9546741

Received 10 January 2017; revised version received 3 August 2017

Appendix: Description of the MyMath intervention procedure

Counting skills
Number rods were used to show that numbers represent quantities and consecutive digits
in the number line have a difference of 1. Addition and subtraction were practised using
number rods. Strategies for quick calculations were presented to students: 1 (e.g.,
5 + 1 = 6, 8 1 = 7), doubles (e.g., 6 + 6 = 12), doubles 1 (e.g., 6 + 7 = 13), +10
(e.g., 10 + 5 = 15), and +9 (e.g., 9 + 5 = 14). Students practised counting by 10 (e.g., 10,
20, 30, . . . and 5, 15, 25, 35, . . .), by 100 (100, 200, 300, . . . and 25, 125, 225, 325, . . .), and
by 110 (3, 113, 223, 333, . . .).

Decomposition strategies
Students were presented with the notion that a given quantity can be formed by
combining other quantities and were allowed to practise with number rods. Emphasis was
given on combinations of 10. Then, they were taught that numbers can be decomposed
into parts in different ways and that they can choose the way that helps them compute
easier. Thus 5 + 7 can be reconceptualized as 5 + (5 + 2), (5 + 5) + 2, and 10 + 2.

Place value
Place value manipulatives were used in order students to understand that digits have
different values in different places within a number. The relation between ones, tens,
20 Elen S. Karakonstantaki et al.

hundreds, and thousands (how ones form tens, hundreds, and thousands, how tens form
hundreds, etc.) was presented. Students practised grouping quantities (e.g., 13 blocks
equal 1 ten-rod and three blocks), representing Arabic numerals as quantities, reading and
writing multidigit numbers up to billions, number comparison, and addition and
subtraction with regrouping (carrying and borrowing).

Solving everyday problems

Problems with currency and time were practised, as they are important in the everyday life
of preadolescent students, and often prove challenging for students with math problems.
They were taught the subdivisions of euro and hour and practised conversions, counting
money using plastic coins and paper money, and telling the time using clocks. They were
also presented with a strategy for problem-solving (Montague, 1998), entailing first
attempting to understand the story and the question, analysing the facts and disregarding
irrelevant information, and finally choosing a strategy, solving the problem and checking
the solution. Students were sometimes presented with false problems, which could not be
solved due to insufficient information, to illustrate that simple calculation of given number
facts is not always sufficient to reach a solution. Although most instruction time was spent
practising with manipulatives and mental computations, and discussing arithmetic
problems, written calculation was also practised and used in solving arithmetic problems.
Having grasped basic number principles, students engaged in written multidigit additions
and subtractions, with and without regrouping.