Research in Developmental Disabilities 72 (2018) 56–66

Contents lists available at ScienceDirect

Research in Developmental Disabilities

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

Research Paper

Are WISC IQ scores in children with mathematical learning MARK

disabilities underestimated? The influence of a specialized
intervention on test performance
⁎
Katharina Lamberta, , Birgit Spinathb
a
Hector Research Institute of Education Sciences and Psychology, University of Tuebingen, Europastrasse 6, 72072 Tübingen, Germany
b
Educational Psychology, Heidelberg University, Hauptstraße 47-51, 69117 Heidelberg, Germany

AR TI CLE I NF O AB S T R A CT

Number of reviews completed is 2 Background: Intelligence measures play a pivotal role in the diagnosis of mathematical learning
Keywords: disabilities (MLD). Probably as a result of math-related material in IQ tests, children with MLD
IQ often display reduced IQ scores. However, it remains unclear whether the effects of math re-
Mathematical learning disabilities mediation extend to IQ scores.
Intervention Aims: The present study investigated the impact of a special remediation program compared to a
Waterglas Intervention Program control group receiving private tutoring (PT) on the WISC IQ scores of children with MLD.
Private tutoring Methods: We included N = 45 MLD children (7–12 years) in a study with a pre- and post-test
WISC
control group design. Children received remediation for two years on average.
Results: The analyses revealed significantly greater improvements in the experimental group on
the Full-Scale IQ, and the Verbal Comprehension, Perceptual Reasoning, and Working Memory
indices, but not Processing Speed, compared to the PT group. Children in the experimental group
showed an average WISC IQ gain of more than ten points.
Conclusion: Results indicate that the WISC IQ scores of MLD children might be underestimated
and that an effective math intervention can improve WISC IQ test performance. Taking limita-
tions into account, we discuss the use of IQ measures more generally for defining MLD in research
and practice.

What the paper adds?

Children with MLD have been found to have lower IQ scores than their typical achieving peers. This might stem from lower
intellectual capacity. However, low IQ test performance might also be the result of poor math skills, since many IQ subtests contain
math-related material. This study is the first to systematically investigate whether improving the math skills of children with MLD
through a specialized intervention program also affects test performance on the WISC IQ test. Understanding the influence of low
math skills on IQ test performance is critical because eligibility for intervention programs and inclusion criteria in scientific research
often partly rely on one-time measurements of IQ scores.
Results indicate that the WISC test performance of children with MLD can be improved via effective remediation programs, and
that math skills seem to be an important factor in WISC performance. The potential underestimation of IQ scores not only calls the
central role of minimum IQ in MLD diagnoses into question but could also explain the conflicting results in research on deficits in
children with MLD. Here, the use of strict and varying IQ cut-off scores as exclusion criteria might result in divergent sample

⁎
Corresponding author.
E-mail addresses: katharina.lambert@uni-tuebingen.de (K. Lambert), birgit.spinath@psychologie.uni-tuebingen.de (B. Spinath).

http://dx.doi.org/10.1016/j.ridd.2017.10.016
Received 22 March 2017; Received in revised form 15 August 2017; Accepted 16 October 2017
0891-4222/ © 2017 Elsevier Ltd. All rights reserved.
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

characteristics, and thus hinder research on MLD core deficits. This study, therefore, provides further evidence in favour of re-
considering the use of minimum IQ cut-offs or discrepancy-based criteria in research and practice.

1. Introduction

Defining and identifying children with mathematical learning disabilities (MLD) is the subject of on-going controversy in the field
of unexpected scholastic underachievement in mathematics. This applies especially to the integral role of IQ, at least in the sense of a
minimum IQ as a marker for MLD. Although there is convincing evidence that children with MLD lack basic numerical skills re-
gardless of the curriculum taught in school, there is little consensus on how to define children with MLD (or developmental dys-
calculia) in contrast to children who simply have math problems (e.g., Kaufmann et al., 2013; Moeller, Fischer, Cress, & Nuerk, 2012).
No core deficit has been identified so far (e.g., Kaufmann et al., 2013; Mazzocco & Myers, 2003), and several authors argue that MLD
is a heterogeneous disorder with several subtypes that might not share the same core deficit (Kaufmann et al., 2013;
Mazzocco & Myers, 2003).
In research and to a certain extent in practice, two formulas are usually used to identify children with MLD. The first is known as
the aptitude-achievement discrepancy criterion (AAD) as outlined by the International Classification of Diseases (ICD-10; World Health
Organisation, 2005). For this, a child’s achievement in math has to be substantially (usually 1–2 SD) lower than their IQ score.
Despite ongoing critical debate on the poor reliability and validity of this measure, the AAD criterion is still used by both psy-
chologists and researchers (cf. Fletcher & Vaughn, 2009; Klassen, Neufeld, & Munro, 2005; Stuebing et al., 2002; Zirkel & Thomas,
2010) in many countries.
The second and now more frequently used formula in research defines MLD as severely low achievement (usually < 10– < 25
percentile on a standardized math test). However, several authors criticize that diagnoses using this formula are highly dependent on
the measure used to test for numerical deficits (e.g., Kaufmann et al., 2013). Desoete, Roeyers, and Clercq (2004) showed that
prevalence rates vary substantially as a function of the measure used.
Moreover, a minimum IQ is required in many cases to differentiate between children with MLD and children with intellectual
disabilities. The threshold for this minimum IQ varies as well, but usually lies between 70 and 85, without taking measurement error
or confidence intervals into account. Often, this minimum level of aptitude is also relevant in AAD diagnoses. As a result of the
variation in these definitions, a child with a one-time IQ measurement of 84 would not be considered math disabled but rather
intellectually disabled in one study (e.g., Brankaer, Ghesquière, & De Smedt, 2014; Schleifer & Landerl, 2011), but would have to have
an IQ score of 69 to be excluded from MLD research or specialized intervention in another (e.g., Fuchs et al., 2009). Several authors
have shown that these varying criteria lead to wide variation in what children are classified as MLD (e.g., Mazzocco & Myers, 2003;
Murphy, Mazzocco, Hanich, & Early, 2007). Moreover, low achievement and AAD criteria do not seem to overlap much, and AAD
criteria appear to be less stable than the former (e.g., Mazzocco & Myers, 2003).
In addition to these problems, children’s concentration and motivation can fluctuate. Overlooking measurement error in math
measures more generally, but especially in IQ test scores, can lead to misdiagnoses in the sense of either false positives or false
negatives, and can also affect the reproducibility and generalizability of findings in learning disabilities research (cf. Cotton,
Crewther, & Crewther, 2005; Dennis et al., 2009).
Furthermore, the empirical basis for differentiating between discrepant (MLD) and non-discrepant (low-achieving) children as
well as between low-achieving children with above (MLD) and below threshold IQ (intellectually disabled) is lacking. There is stable
evidence from reading disabilities research that these groups cannot be differentiated on the basis of any of a number of reading-
related and cognitive measures, and no differences in response to intervention have been found (cf. Stuebing et al., 2002). A smaller
number of similar results have been found for children with MLD. For example, Jiménez Gonzalez and Espinsel (1999) demonstrated
that measures of core cognitive characteristics do not differ among children with math difficulties and a below-average IQ and
children with an IQ within the normal range. Furthermore, math disabled discrepant children do not differ from non-discrepant
children on word problems (Jiménez Gonzalez & Espinsel, 1999), arithmetic fact retrieval (Busch, Oranu, Schmidt, & Grube, 2013),
numerical magnitude processing (Brankaer et al., 2014), or working memory capacities (e.g., Maehler & Schuchardt, 2011).
Furthermore, IQ was not found to predict the growth rates of various mathematical competencies such as exact calculation, place-
value tasks, story problems or approximate calculation (e.g., Jordan, Hanich, & Kaplan, 2003). There are some contradictory findings
showing differences between groups (e.g., Tolar, Fuchs, Fletcher, Fuchs, & Hamlett, 2016); however, the severity of impairment
seems to be the decisive factor in explaining differences in cognitive variables (c.f. Elliott & Resing, 2015; Tolar et al., 2016).
In reaction to the ongoing reliability and validity critiques of the role of IQ in the defining and identifying learning disabilities
(Büttner & Hasselhorn, 2011; Mazzocco & Myers, 2003; Stanovich, 2005), some countries – like the United States – have shifted to a
multi-tiered approach applying response-to-intervention criteria (RtI). Here, children are classified as learning disabled if they do not
significantly improve as a result of remediation. However, this approach is seldom used in MLD research, probably for economic
reasons. In addition, in many countries, the implementation of systematic multi-tiered intervention is still in its early stages (if it
exists at all), and RtI criteria are not feasible. As a result, access to and eligibility for MLD remediation programs or special education
support often require AAD or at least severely low achievement, often in combination with a minimum IQ criterion, which can lead to
delayed access to these programs in some cases (c.f. Stuebing et al., 2002). This is especially true in many European countries, where
multi-tiered intervention programs have not been systematically implemented in public schools. Some countries, like Belgium, have
combined different approaches to diagnosing MLD to improve consistency (e.g., Desoete et al., 2004); however, IQ is still an essential
component of the diagnosis when AAD or minimum IQ criteria are used.
For this reason, further investigations of IQ scores among children with MLD is of great importance for research as well as for

57
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

affected children and parents. It has been argued that reading deficits can push down IQ scores (e.g., Stuebing et al., 2002), thus
reducing students’ chances of satisfying the minimum IQ or AAD criteria and rendering them ineligible for intervention. The same can
be assumed for math deficits. Many common intelligence tests include tasks requiring skills that are assumed to be deficient in
children with MLD. For example, some IQ subtests use numerical material or even require basic arithmetic competences (e.g., K-ABC,
Melchers, Preuß, & Kaufman, 2007; WISC-IV, Wechsler, 2003), which are obviously impaired in children with MLD. For the Wechsler
Intelligence Scale for Children (WISC-III and -IV), which was used in the current study, this particularly applies to Working Memory
Index subtests, where studies have shown that children with MLD display problems compared to typically achieving children (e.g.,
Geary, Hamson, & Hoard, 2000; Swanson & Sachse-Lee, 2001). Furthermore, many tests include subtests that require abilities con-
sidered to be associated with math skills, such as visuospatial perception, which has been shown to be inferior in children with MLD
compared to typically achieving children (e.g., Lambert & Spinath, 2017; Reuhkala, 2001).
In addition, children with MLD show deficits in problem solving tasks (e.g., Fuchs, Fuchs, Stuebing et al., 2008; Swanson,
Lussier, & Orosco, 2015), which might affect subtests requiring complex problem solving. As a consequence, the structure of IQ tests
might lead to underestimations of the intellectual potential of children with MLD, while at the same time over-identifying high-IQ
children (Callinan, Cunningham, & Theiler, 2013).
Indeed, in a striking number of studies, the mean IQ scores of children with MLD are substantially lower than those of children
with typical achievement or even children with reading or spelling disabilities (e.g., Fuchs, Fuchs, Craddock et al., 2008; Geary,
Hoard, Nugent, & Bailey, 2012; Jordan et al., 2003; Poletti, 2016). IQ scores are also thought to be influenced by learning en-
vironments and products of learning (c.f. Elliott & Resing, 2015), which are impaired in children with MLD. This raises the question of
whether special interventions that target MLD could also alter IQ scores. If MLD weakens affected children’s performance on IQ test
items requiring skills that are typically impaired in such children, intervention programs that improve math skills might in turn lead
to better outcomes on IQ tests. However, only a few studies have addressed this issue. Dowker and Sigley (2010) found that targeted
math intervention improves the performance of children with MLD in the Arithmetic subtest of the WISC-III. This is not surprising, as
this subtest involves several calculation tasks. White (1988) conducted a meta-analysis on the effects of direct instruction on special
education students’ achievement outcomes, drawing on 25 studies examining direct instruction treatments for children experiencing
learning problems such as a learning disability, reading disability, emotional disability, or mental retardation. In addition to findings
that direct instruction had positive effects on academic outcomes compared to controls, the study-weighted mean ES was 1.32 for
intellectual ability. However, according to White, these findings should be interpreted with caution due to the small number of
studies (3) including intellectual ability measures. More recently, Fischer-Klein (2007) analyzed data from 20 children with MLD in
her dissertation. Of those, ten children participated in an unevaluated intervention program (“Therapieprogramm Mathematik
Wien”), while the other ten did not receive any specific training for 5 months. Children in the intervention group exhibited a
significant increase in average math achievement as well as a 6.9-point increase in nonverbal IQ, whereas no increases were reported
for the control group. As with White (White, 1988), Fischer-Klein’s (Fischer-Klein, 2007) results have to be treated with caution due
to methodological problems such as small sample sizes and having the same person conduct training and testing.
Due to these limitations and the lack of additional studies, reliable statements about changes in IQ scores as a result of MLD
interventions cannot yet be made. The vast majority of intervention studies gather IQ data at one measurement point only. As a result
of this, as well as the integral importance of IQ in MLD research and practice, the present study sought to investigate the effects of
MLD intervention on the WISC IQ scores (Wechsler Intelligence Scale for Children) of children with MLD, thus contributing to the
discussion of IQ scores as a marker in the definition of MLD.

1.1. Objectives

The current study investigated whether a special intervention program for children of all ages with MLD would affect their WISC
IQ scores compared to a control group. The WISC is one of the most popular IQ tests used by school psychologists in many countries to
assess aptitude in children with learning difficulties (Evers et al., 2012). It allows a general IQ score to be determined while also
enabling different aspects of children’s abilities to be assessed.
The experimental group received support through the Waterglass Intervention Program (WIP). The WIP is a program designed for
children with MLD of all ages and class levels and represents a form of extracurricular MLD training in Germany. In this program,
basic mathematics skills are taught systematically, with curriculum-based elements gradually implemented until a child has caught
up with his or her class. Establishing basic math skills has been found to be the most effective treatment for children with MLD (cf.
Kroesbergen & Van Luit, 2003) and is a necessary prerequisite for learning higher-order mathematics (Mercer and Miller, 1992; Van
Luit and Naglieri, 1999).
The control group, in contrast, received private tutoring in mathematics. In Germany, PT usually does not address these issues in
depth but rather relies on teaching curricular facts, giving feedback and tips, and asking questions and choosing exercises (Wittwer,
2008). It has been shown that children with MLD do not benefit from PT over time, especially when compared with the WIP
(Lambert & Spinath, 2014). After two years, more than two-thirds of children enrolled in the WIP exhibited an achievement level
within the normal range, whereas only 10% of the tutored children did so. Most of the tutored children still numbered among the
lowest 15% of students in their grade level. In addition, children who participated in the WIP showed significantly greater im-
provements on psychological measures such as test anxiety, school aversion, and depressive symptoms (Lambert & Spinath, 2013). On
the basis of these results, we predicted that PT would not be sufficient to effectively promote IQ scores in MLD children either.
Building on the fact that the WISC includes subtests with math-dependent or math-related material as well as the fact that MLD
children tend to have lower IQ scores than their peers with typical achievement, the following hypotheses were evaluated. We

58
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

Table 1
Demographic Data at Program Entry by Group.

EG (n = 25) CG (n = 20)

Gender Gender
male 10 male 5
female 16 female 15
total 25 total 20

Age (months) Age (months)

mean 106.27 mean 101.35
SD 16.70 SD 8.20
range 87–149 range 86–119

SES (HISEI) SES (HISEI)

mean 58.38 mean 55.40
SD 15.83 SD 15.70
range 29–88 range 24–88

Duration of intervention (months) Duration of intervention (months)

mean 22.46 mean 23.70
SD 7.80 SD 8.39
range 10–38 range 11–37

Note: EG = Experimental group, CG = Control Group, SES = Socio-economic Status.

hypothesized that children enrolled in the WIP, a specialized MLD training program, would show significant WISC IQ score gains over
time and that these improvements would be higher than in the PT group. In addition, we expected that the higher WISC IQ score gains
in the WIP group would be the result of gains on subtests related to math difficulties (Working Memory and Perceptual Reasoning)
rather than subtests assessing verbal abilities or processing speed.

2. Methods

2.1. Participants

The sample consisted of N = 45 children between the ages of 7 and 12 diagnosed with MLD according to ICD-10 criteria (the
sample was identical to the sample analyzed in Lambert & Spinath 2013; Lambert & Spinath2014). Children were classified as MLD
when (1) their math achievement was below the 16th percentile, (2) but they had an IQ > 80, (3) had not been diagnosed with any
clinically relevant psychological or neurological disorder, and (4) did not meet the requirements for special education status (isolated
learning disabilities do not make one eligible for special education status in Germany). All children were Caucasian, and two children
had one parent born outside Germany (France and Turkey). No children in the sample reported having reading disabilities.
The sample was not random, because the intervention centers had different numbers of slots available, and children’s parents had
to pay for both interventions. However, parents were able to apply for financial support from youth welfare services. All parents
received an offer to enroll their children in the WIP.1
Children entered the program during elementary school or lower secondary school, ranging from Grades 2 to 6. As can be seen in
Table 1, the groups did not differ on any measure or distribution.

2.2. Procedures

All children received one training session a week during the school year. Each session lasted 50–60 min and took place at one of
the WIP or PT facilities participating in the study. The WIP was conducted by seven specially trained psychologists and early
childhood teachers. The program starts with the base-10 structure of our number system and gradually builds up basic numerical
skills according to the developmental model by Krajewski and Schneider (2009). Only afterwards are the main content areas of
school-based mathematics taught. The program is highly individualized and includes the use of a special form of manipulatives to
display mathematical concepts and procedures. Children are encouraged to relate the mathematical content to the material and
reflect their actions verbally. For a more detailed description, see Lambert and Spinath (2014).
Children in the control group received an equal amount of PT conducted by math teachers in private facilities specializing in
extracurricular PT. Extracurricular PT is common among German students. As mentioned above, the content of the PT depended on
the child’s school curriculum. The children’s schoolbooks formed the basis of the lessons, and additional worksheets were also
provided. There was no special program or teaching method. Children usually reviewed their schoolwork or studied for exams.

1
We rejected the option of including an untreated control group since it would not have been justifiable to impede children with severe problems in mathematics
from getting support for a two-year period.

59
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

Fig. 1. Composition of subtests used to calculate the WISC IQ index scores and the FSIQ in the present study.

2.3. Measures

2.3.1. Intelligence
Intelligence was measured with the German version (HAWIK-IV: Petermann & Petermann, 2007; HAWIK-III; Tewes,
Rossmann, & Schallberger, 2002) of the Wechsler Intelligence Scale for Children – Editions 3 and 4 (WISC-III: Wechsler, 1991; WISC-
IV: Wechsler, 2003).
Funding requirements from Germany’s youth welfare services forced us to switch to the newly released WISC-IV after the study
was already in progress. However, we ensured that the percentage of children completing versions three and four did not differ
between groups. 70% (CG) and 73% (EG) completed the WISC-III, while 30% (CG) and 27% (EG) completed the WISC-IV. The same
version was used before and after the intervention so that changes were not made in the middle of testing any individual child.
Despite some changes to the make-up of subtests and tasks (see Fig. 1), Lipsius and Petermann (2009) demonstrated that the two
versions are correlated with r = 0.87. Whenever possible, corresponding subtests in WISC-III and WISC-IV were used (see Fig. 1) to
better compare results obtained using the two versions. On average, the WISC-III produced an IQ score that was 3 points higher. In
this sample, there were no significant differences in IQ scores measured by the WISC-III and WISC-IV (T0: t(43) = 0.02, p = 0.99; T1:
t(43) = 0.04, p = 0.97), and the results did not change after excluding all children who completed the WISC-IV.
Both editions of the WISC generate a Full-Scale IQ (FSIQ); in addition, we computed four indices representing different cognitive
abilities: the Verbal Comprehension Index (VCI), Perceptual Reasoning Index (PRI), Processing Speed Index (PSI), and Working
Memory Index (WMI). The index labels differed slightly between editions; we use the labels from the WISC-IV in this paper. On the
WISC-III, the FSIQ could additionally be subdivided into a verbal (VIQ) and a perceptual IQ (PIQ), but not this was not the case for the
WISC-IV. Fig. 1 displays all subtests used in this study to determine the FSIQ and presents all indices separately for the two editions.
To determine the IQ scores, the raw data were first converted into subtest scores ranging from 1 to 19 (M = 10, SD = 2). These scores
were then added together and converted into IQ scores on the basis of the applicable norm tables.

3. Results

3.1. Effects of the intervention on Full-Scale IQ and changes in indices

All means, standard deviations, and significance tests are displayed in Table 2. The experimental group’s mean IQ at the beginning
of the intervention was 94.68 (SD = 8.75). The control group demonstrated an average cognitive ability of 93.80 (SD = 10.60). The
groups did not differ on the FSIQ or on any other index.
A repeated-measures ANOVA was computed to test for effects of the intervention on the IQ measures. In line with our predictions,
the analyses revealed significant main effects for time and group, as well as a significant time x group interaction effect, which
indicated a different pattern of change over time favoring the experimental group. Whereas the experimental group’s IQ scores

60
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

Table 2
Means, Standard Deviations, and Statistical Information for the WISC Full-Scale IQ and all Indices

T0 T1

EG (n = 25) CG (n = 20) EG (n = 25) CG (n = 20)

M SD M SD M SD M SD F df p η2

FSIQ T 40.22 < 0.001 0.48

94.68 8.75 93.80 10.66 105.84 9.55 95.20 11.26 G 4.14 1, 43 < 0.05 0.09
TxG 24.28 < 0.001 0.36
VCI T 8.65 < 0.01 0.17
100.16 11.49 98.85 9.54 108.36 12.05 97.60 11.08 G 3.71 1, 43 0.06
TxG 15.99 < 0.001 0.27
PRI T 49.92 < 0.001 0.54
90.68 9.31 89.75 10.25 104.20 8.32 92.75 11.89 G 5.15 1, 43 < 0.05 0.11
TxG 20.24 < 0.001 0.32
WMI T 2.15 0.15
93.04 8.02 96.25 11.42 100.88 10.46 93.05 10.66 G 0.80 1, 43 0.34
TxG 12.20 < 0.01 0.22
PSI T 7.39 < 0.01 0.15
100.80 10.85 102.35 12.68 105.28 8.58 104.45 12.65 G 0.01 1, 43 0.91
TxG 0.97 0.33
VIQa T 4.12 0.05 0.12
97.68 9.56 97.79 9.62 104.11 10.46 95.57 12.06 G 1.44 1, 31 0.24
TxG 17.37 < 0.001 0.36
PIQa T 36.60 < 0.001 0.54
89.63 8.36 93.71 13.75 104.16 8.68 98.43 13.29 G 0.06 1, 31 0.82
TxG 9.52 < 0.01 0.24

Note: EG = Experimental group, CG = Control Group, FSIQ = Full-Scale IQ, VCI = Verbal Comprehension Index, PRI = Perceptual Reasoning Index,
PSI = Processing Speed Index, WMI = Working Memory Index. The significance is indicated by the p value, p < 0.001 = ***, p < 0.01 = **, p < 0.05 = *
a
WISC-III only.

improved by more than 10 IQ points (M = 11.16), the PT group gained only 1.40 IQ points.
In a second step, a repeated-measures MANOVA was used to determine the effects of the intervention on the four WISC indices.
Analyses with the VCI, PRI, PSI, and WMI as within-subjects variables and group as the between-subjects variable showed a main
effect of time, F(1, 43) = 13.09, p < 0.001, η2 = 0.57, but not of group, F(1, 43) = 1.91, p = 0.13. However, the interaction was
significant, F(1, 43) = 8.28, p < 0.001, η2 = 0.45, indicating different patterns of change in the groups over time.
A univariate analysis of variance with repeated measures was applied to test for differences in the indices (see Table 2 for results).
In line with our predictions, significant interaction effects were found for the PRI and WMI, with middle to high effect sizes in favor of
the experimental group. The VCI index also showed a significant interaction, once again due to a larger increase in the experimental
group’s scores (increase of 8.20 points) compared to the control group (decrease of 1.25 points). For the PSI, there was a change over
time only; the groups did not differ over the course of the intervention.

3.2. Subtest changes

FSIQ changes might have been due to improvements in certain subtests associated with math disabilities, such as the Arithmetic,
Block Design, or Digit Span subtests. Therefore, separate repeated-measures ANOVAs were computed for each subtest because not all
children completed all subtests.
As can be seen in Tables 3 and 4, for the WMI, a significant time by group interaction effect was found for the Digit Span subtest,
and a marginally significant interaction was found for the Arithmetic subtest. Thus, we found the anticipated interaction in favor of
the experimental group on the WMI, but only in one subtest. The analyses also revealed significant or marginally significant in-
teraction effects for the PRI Picture Completion and Block Design subtests. Contrary to our expectations, there were significant or
marginally significant changes over time favoring the experimental group on all VCI subtests. In line with our predictions, there were
no significant interaction effects on the PSI, and only one main effect of time for the Coding subtest.
The PRI subtests differed slightly across the subgroups of the sample. Whereas all children completed the Block Design and Picture
Completion subtests, only n = 33 children completed the Picture Arrangement and Object Assembly subtests. A significant inter-
action was found for Picture Arrangement, but not for Object Assembly (Table 4). Due to the small sample size of children completing
the WISC-IV Matrix Reasoning and Picture Concepts subtests, only means and standard deviations are reported.
In summary, in line with our predictions, there were significant changes over time on some PRI and WMI subtests in favor of the
experimental group.

61
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

Table 3
Means, Standard Deviations, and Statistical Information for the Verbal Comprehension, Working Memory, and Processing Speed Index Subtests.

T0 T1

EG (n = 25) CG (n = 20) EG (n = 25) CG (n = 20)

M SD M SD M SD M SD F df p η2

Comprehension T 0.41 0.52

10.00 2.96 10.65 1.79 11.00 1.53 10.10 2.40 G 0.05 1, 43 0.83
TxG 4.89 < 0.05 0.10
Similarities T 9.84 < 0.01 0.19
9.36 2.36 10.10 2.08 11.40 2.20 10.00 2.13 G 0.32 1, 43 0.58
TxG 11.97 < 0.01 0.22
Vocabulary T 0.16 0.69
9.68 2.73 9.50 2.19 10.60 2.48 8.90 2.10 G 0.24 1, 43 0.13
TxG 3.62 0.06
Information T 1.24 0.27
9.80 2.87 9.45 2.06 11.12 2.33 8.95 2.44 G 3.86 1, 43 0.06
TxG 6.12 < 0.05 0.13
Arithmetic T 1.58 0.22
8.32 2.46 8.35 1.90 9.56 2.34 8.10 2.00 G 1.87 1, 43 0.18
TxG 3.58 0.07
Digit Span T 0.59 0.45
9.24 2.17 10.20 2.86 10.32 2.58 9.70 1.81 G 0.08 1, 43 0.70
TxG 4.40 < 0.05 0.09
Coding T 10.80 < 0.01 0.20
10.00 2.31 10.20 2.76 11.48 1.71 10.95 3.14 G 0.06 1, 43 0.81
TxG 1.16 0.29
Symbol Search T 0.00 0.95
10.16 2.76 10.70 2.16 10.44 2.16 10.45 2.14 G 0.18 1, 43 0.68
TxG 1.15 0.29

Note: EG = Experimental group, CG = Control Group. The significance is indicated by the p value, p < 0.001 = ***, p < 0.01 = **, p < 0.05 = *

Table 4
Means, Standard Deviations, and Statistical Information for the Perceptual Reasoning Index Subtest.

T0 T1

EG (n = 25) CG (n = 20) EG (n = 25) CG (n = 20)

M SD M SD M SD M SD F df p η2

Block Design T 10.50 < 0.01 0.20

8.64 2.22 8.05 2.61 10.32 1.82 8.50 2.28 G 4.30 1, 43 < 0.05 0.09
TxG 3.50 0.07
Picture Completion T 11.83 < 0.001 0.22
7.64 2.77 8.25 2.15 10.08 1.75 8.75 2.67 G 0.40 1, 43 0.53
TxG 5.15 < 0.05 0.11
Picture Arrangementa T 14.70 < 0.001 0.30
9.00 2.43 8.56 3.01 11.10 2.38 9.13 2.90 G 2.15 1, 34 0.15
TxG 4.90 < 0.05 0.15
Object Assemblya T 2.62 0.12
9.15 2.76 8.94 2.72 10.05 1.73 9.19 3.17 G 0.45 1, 34 0.51
TxG 0.84 0.37
Picture Conceptsb 9.17 1.33 8.56 3.28 12.67 2.16 9.33 2.24
Matrix Reasoningb 8.83 2.56 9.00 3.91 11.33 1.87 9.11 2.85

Note: EG = Experimental group, CG = Control Group. The significance is indicated by the p value, p < 0.001 = ***, p < 0.01 = **, p < 0.05 = *
a
EG: n = 20; CG: n = 16.
b
EG: n = 6; CG: n = 9.

4. Discussion

In many countries, IQ measures are of crucial importance in diagnosing mathematical learning disabilities, and are often used to
identify children with MLD in research. Children with severe math difficulties who score above a given IQ threshold are considered to
have MLD (low achievement criterion), whereas children who miss the cut-off by even one point are either categorized as low-
achieving or intellectually disabled. In many cases, these children are excluded from research and in some countries even from
specialized MLD intervention programs. The standard AAD diagnostic method is also still used, albeit to a lesser extent, in research

62
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

and practice to classify students as having mathematical learning disabilities – despite the fact that this method lacks reliability and
validity (e.g., Büttner & Hasselhorn, 2011; Dombrowski and Gischlar, 2014) and despite its association with delays in accessing
remediation programs (Stuebing et al., 2002). However, hardly any studies have evaluated whether the effects of specialized math
interventions can improve the WISC IQ scores of MLD children. The data presented here add to the literature concerning IQ scores’
lack of validity for diagnosing MLD. We hypothesized that the WISC IQ scores of MLD children might underestimate their intellectual
potential and could be positively affected by a specialized intervention program. The results of the present study support this pre-
diction and are in line with a small number of previous studies showing that the IQ scores of children with learning disabilities can
change as a result of specialized training (Fischer-Klein, 2007; White, 1988). The WIP (Schlotmann, 2004), the specialized inter-
vention program applied in this study, has previously been shown to produce greater improvements in MLD children than PT. In the
current study, children participating in the MLD intervention program gained more than ten points on the FSIQ on average, whereas
children in the control group who received PT did not improve significantly. These results indicate that a specialized intervention
program can improve the WISC IQ scores of children with MLD. This does not imply that these children become more intelligent, but
rather that mathematics skills seem to be an important factor in WISC performance, calling into question the use of the WISC in the
MLD diagnostic process. Similarly, Giofrè and Cornoldi (2015) proved that children with learning disabilities in general display a
different factorial structure on the WISC than typical achieving children and that the WISC should therefore not be used with these
children. Similar studies of other IQ tests such as the Raven Progressive Matrices (Raven, Court, & Raven, 1984) are still lacking.
However, the findings of Fischer-Klein (2007), who used a figural IQ test to assess fluid intelligence, indicate that the results of the
present study can be generalized to different IQ tests, thus calling into question their use in MLD diagnostics. However, future studies
need to clarify whether certain IQ tests might not disadvantage this group of children.
The results of the present study imply a transfer effect of the MLD intervention, since this training was shown to produce larger
gains in math skills over time compared to PT (Lambert & Spinath, 2014). Furthermore, it can be concluded that the WISC IQ scores of
MLD children might often be underestimated, resulting in a lower likelihood that such children will be diagnosed with MLD. This is
especially true for children with WISC IQ scores at the lower end of the IQ scale, whereas high-IQ children might be over-identified
(Callinan et al., 2013). As a result, children with MLD might not receive appropriate support or face a delay before being accepted
into a remediation program, thus possibly leading to more severe deficits. In the case of low-IQ children, who might not be able to
compensate as much as high-IQ children, this delay might result in even more severe disadvantages over time. As expected, increases
in the WIP group’s WISC IQ scores were due to higher gains on the PRI subtests and WMI indices in particular, but not on the PSI.
Most WMI subtests on the WISC encompass primarily numerical material. The present findings indicate that intervention programs –
such as the WIP (Lambert & Spinath, 2014) – that successfully improve MLD children’s math skills also lead to improvements in WM
measures involving numerical material. This is in line with previous findings that MLD children exhibit particularly low WM skills
when the measures contain numerical material (De Weerdt, Desoete, & Roeyers, 2013; Passolunghi & Cornoldi, 2008). The present
study demonstrates that WM skills assessed using such measures might also be underestimated and that improvements in math
abilities can be transferred to specific WM tasks.
The largest effect sizes for the time by group interaction were found for the PRI. The PRI subtests are supposed to examine
nonverbal fluid intelligence, including visual perception and visuospatial processing. Impairments in these skills are often associated
with MLD (e.g., Jordan & Hanich, 2003). The present study underpins this notion.
Contrary to our expectations, children in the experimental group also exhibited substantial improvements on most VCI subtests
even though they include no obvious math-related content. There are several possible explanations for this result. First, the present
findings might provide further evidence that some arithmetic skills are – at least in part – language-dependent or associated with
language proficiency (e.g., Gelman and Gallistel, 2004; von Aster and Shalev, 2007). It has been shown that low reading compre-
hension skills predict subsequent performance on word problem-solving skills among primary school (Bjork and Bowyer-Crane, 2013;
Pimperton and Nation, 2010) and secondary school students (Björn, Aunola, & Nurmi, 2016). Decoding skills and phonological
awareness have also been found to affect specific math skills (Bjork & Bowyer-Crane, 2013; Jordan et al., 2003). Activation patterns
in fMRI data while participants solved simple memorized arithmetic tasks underscore the relationship between literacy and math
skills (Dehaene, Spelke, Pinel, Stanescu, & Tsivkin, 1999). In addition, children with language impairments tend to have comorbid
difficulties in acquiring math skills (Rousselle, Palmers, & Noël, 2004). In the present sample, however, mean scores on the VCI were
within the normal range at intervention onset, and no child with severe reading disabilities were included. Therefore, it can be
assumed that the WIP group’s improving math skills were accompanied by or influenced improvements in the verbal reasoning
abilities assessed on the WISC VCI subtests. However, the mechanisms underlying this relationship remain unclear in the present data
and should be addressed in future research.
Moreover, both these results as well as the improvement on the PSI scale might be explained by the method used in the WIP.
Children were highly encouraged to verbally reflect on the actions they performed with the manipulatives and translate them into
abstract mathematical notation. In this way, children might indirectly learned comprehensive problem solving skills, which are also
often impaired in children with MLD (e.g., Fuchs & Fuchs, 2002). Therefore, a math intervention might specifically affect verbal
reasoning and problem solving skills, but not verbal skills in general.
Second, the WIP group’s increases in VCI might be explained by the intervention’s emphasis on verbal reflection on calculation
processes. WIP trainers may have encouraged children to use verbal reflection, thereby fostering the children’s verbal reasoning
skills. Third, an alternative explanation for changes in VCI scores might be that the WIP did not influence language-related skills
directly but rather served (in part) as a cognitive training that was able to positively affect WISC IQ scores and math text problem
solving skills (Klauer & Phye, 2008; Sonntag, 2004). Additional research is needed to clarify whether children with MLD’s WISC IQ
scores, and IQ scores in general, can be equally improved by cognitive trainings and math-specific interventions, or whether gains in

63
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

IQ scores require the improvement of math skills first or simultaneously.

The present study had some limitations in terms of its small sample size and lack of randomization. As we did not analyze a
representative sample, it cannot be fully ruled out that our findings are not fully transferable to all children with MLD. Parents had to
actively enroll their children in one of the programs. This might have caused a biased sample towards a slightly higher socioeconomic
status. However, the range of HISEI in the present sample does not indicate bias in this direction, and any biases that would exist
would apply to both groups. Since the children who received PT did not significantly improve their IQ scores, it can be concluded that
the determining factor in IQ score gains was the remediation of math skills, and this finding should also transfer to different samples.
Regarding the lack of randomization, it cannot be ruled out that the samples differed on variables other than those assessed. For
example, the parents of children participating in the WIP might have put more emphasis on their children’s improvement and
practiced more outside of the training sessions. These aspects might have, in part, caused the large effect sizes. However, greater
ambition and more practice with parents and others might also have influenced initial math skills. It is possible that bias in this
direction might have led to more drill training or practicing more tasks. However, it does not seem plausible that parents differed in
terms of teaching the solution strategies relevant for IQ tests. As a result, the findings of the present study do not necessarily show that
the WIP itself improves IQ scores but rather that this increase can be attributed to gains in math skills. Therefore, the results should
also generalize to other intervention programs and randomized samples in which math skills improve.

4.1. Implications and conclusion

Despite these limitations, the results of the present study indicate that the WISC IQ scores of MLD children might be under-
estimated in many cases because they have trouble in math. Our findings suggest that the IQ scores of MLD children in general can
undergo substantial changes with specialized interventions. Certainly, this assumption should be treated with caution due to con-
straints concerning the design of the study. Thus, we cannot rule out the possibility that factors other than the type of intervention
were accountable for the group differences. However, even if these group differences were due to other factors, the present findings
call into question the validity of the integral role of IQ scores in MLD research and practice by demonstrating MLD children’s IQ
scores. In addition, it has to be taken into account that the results of studies excluding children with low IQs might not be re-
presentative of all children with MLD.
The present results further call into question the use of discrepancy-based or minimum IQ criteria for MLD identification.
However, what do these results imply for other diagnostic criteria? Within the RtI framework, it could be argued that IQ scores as a
whole are irrelevant for eligibility for remediation programs. However, in practice, even here, minimum IQ cut-off scores are reg-
ularly used to distinguish children with learning disabilities from those with intellectual disabilities. The present study questions
whether such a clear distinction is possible before children have participated in a math-specific remediation program (and maybe
even afterwards). In addition, apart from practical problems related to the lack of a systematic multi-tiered remediation system in
many countries, it has been argued that the RtI approach alone is not sufficient and reliable for identifying children with learning
disabilities (for an overview see Hale et al., 2010).
Several authors have proposed a “third approach” to diagnosing learning disabilities: patterns of strengths and weaknesses (PSW).
This approach requires an individual discrepancy between cognitive or processing constructs related to the academic deficit, aca-
demic performance and unrelated cognitive abilities (c.f. Flanagan, Fiorello, & Ortiz, 2010). For the present study, it could be argued
that children with MLD exhibit a specific pattern of weaknesses in the WMI and PRI compared to the other indices at intervention
onset. However, many studies question the stability and reliability of classifications based on PSW (e.g., McGill, Styck,
Palomares, & Hass, 2015; Taylor, Miciak, Fletcher, & Francis, 2017; Watkins, 2000). In the present study, the pattern mentioned
above only applied on an average basis and can therefore not be generalized to any individual child. In addition, the PSW framework
would have predicted transfer effects on the WMI and PRI alluding to the presence of MLD. However, gains were only found for the
WMI, as well as the VCI, which cannot be explained by the PSW framework.
Taken together, the present results do not provide support for any of the criteria used to classify children as MLD. Despite the
problems exhibited by IQ measures in the MLD diagnostic process, they can still provide valuable information about specific char-
acteristics of a child’s individual cognitive pattern and should therefore not be fully ignored. However, the present results strongly
underline the findings of simulation studies that the use of fixed cut-off points – which more or less apply to every single criterion –
are unreliable at the individual level, especially when it comes to IQ measures (e.g., Burns, Scholin, Kosciolek, & Livingston, 2010;
Francis et al., 2005; Miciak, Taylor, Denton, & Fletcher, 2015). Instead of providing evidence for a specific type of criteria, the present
results imply that IQ scores of children with MLD should be treated with caution
There is an urgent need to replicate the results of this study in larger and randomized samples as well as using different kinds of IQ
measures. A confirmation of our findings will indicate the need to reconsider the requirement by many countries and school boards
that children have an IQ of at least 80 or 85 to be eligible for MLD-specific remediation irrespective of other diagnostic criteria, the
use of AAD criteria in determining eligibility for interventions, and the common practice of excluding children with low IQs from
studies on MLD (Fuchs & Fuchs, 2002; Kaufmann et al., 2003). Mazzocco and Myers (2003) showed that research results and con-
clusions differ with respect to the criteria and cut-offs used in MLD studies. Therefore, excluding such children not only affects their
personal life and career options, but might also distort research on MLD children’s core deficits.

Acknowledgements

We wish to thank all participating children and their parents as well as the teachers involved for their benevolent cooperation.

64
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

Part of this research was supported by the Postdoc Academy of the Hector Research Institute of Education Sciences and
Psychology, Tübingen, funded by the Baden-Württemberg Ministry of Science, Education and the Arts. Katharina Lambert is also
member of the LEAD Graduate School [GSC1028], funded by the Excellence Initiative of the German federal and state governments.

References

Bjork, I. M., & Bowyer-Crane, C. (2013). Cognitive skills used to solve mathematical word problems and numerical operations: A study of 6- to 7-year-old children.
European Journal of Psychology of Education, 28(4), 1345–1360.
Björn, P. M., Aunola, K., & Nurmi, J.-E. (2016). Primary school text comprehension predicts mathematical word problem-solving skills in secondary school. Educational
Psychology, 36(2), 362–377. http://dx.doi.org/10.1080/01443410.2014.992392.
Brankaer, C., Ghesquière, P., & De Smedt, B. (2014). Numerical magnitude processing deficits in children with mathematical difficulties are independent of in-
telligence. Research in Developmental Disabilities, 35(11), 2603–2613. http://dx.doi.org/10.1016/j.ridd.2014.06.022.
Burns, M. K., Scholin, S. E., Kosciolek, S., & Livingston, J. (2010). Reliability of decision-making frameworks for response to intervention for reading. Journal of
Psychoeducational Assessment, 28(2), 102–114. http://dx.doi.org/10.1177/0734282909342374.
Busch, J., Oranu, N., Schmidt, C., & Grube, D. (2013). Rechenschwäche im Grundschulalter: Reduzierte Verfügbarkeit basalen arithmetischen Faktenwissens und
Belasung des Arbeitsgedächtnisses bei Drittklässlern. [Arithmetic disabilities in elementary school years: Difficulties in retrieving basic arithmetic facts and
working memory load in third graders]. Lernen und Lernstörungen, 2(4), 217–227.
Büttner, G., & Hasselhorn, M. (2011). Learning disabilities: Debates on definitions, causes, subtypes and responses. International Journal of Disability Development and
Education, 58(1), 75–87. http://dx.doi.org/10.1080/1034912X.2011.548476.
Callinan, S., Cunningham, E., & Theiler, S. (2013). Revisiting discrepancy theory in learning disabilities: What went wrong and why we should go back. Australian
Journal of Guidance and Counselling, 23(1), 1–17. http://dx.doi.org/10.1017/jgc.2012.22.
Cotton, S. M., Crewther, D. P., & Crewther, S. G. (2005). Measurement error: Implications for diagnosis and discrepancy models of developmental dyslexia. Dyslexia: An
International Journal of Research and Practice, 11(3), 186–202. http://dx.doi.org/10.1002/dys.298.
De Weerdt, F., Desoete, A., & Roeyers, H. (2013). Working memory in children with reading disabilities and/or mathematical disabilities. Journal of Learning
Disabilities, 46(5), 461–472. http://dx.doi.org/10.1177/0022219412455238.
Dehaene, S., Spelke, E., Pinel, P., Stanescu, R., & Tsivkin, S. (1999). Sources of mathematical thinking: Behavioral and brain-imaging evidence. Science, 284(5416),
970–974. http://dx.doi.org/10.1126/science.284.5416.970.
Dennis, M., Francis, D. J., Cirino, P. T., Schachar, R., Barnes, M. A., & Fletcher, J. M. (2009). Why IQ is not a covariate in cognitive studies of neurodevelopmental
disorders. Journal of the International Neuropsychological Society, 15(3), 331–343. http://dx.doi.org/10.1017/S1355617709090481.
Desoete, A., Roeyers, H., & Clercq, A. D. (2004). Children with mathematics learning disabilities in Belgium. Journal of Learning Disabilities, 37(1), 50–61. http://dx.doi.
org/10.1177/00222194040370010601.
Dombrowski, S. C., & Gischlar, K. L. (2014). Ethical and empirical considerations in the identification of learning disabilities. Journal of Applied School Psychology, 30,
68–82. http://dx.doi.org/10.1080/15377903.2013.869786.
Dowker, A., & Sigley, G. (2010). Targeted interventions for children with arithmetical difficulties. BJEP monograph series II, number 7 – understanding number development and
difficulties65–81. http://dx.doi.org/10.1348/97818543370009X12583699332492.
Elliott, J. G., & Resing, W. C. M. (2015). Can intelligence testing inform educational intervention for children with reading disability? Journal of Intelligence, 3,
137–157. http://dx.doi.org/10.3390/jintelligence3040137.
Evers, A., Muñiz, J., Bartram, D., Boben, D., Egeland, J., Fernández-Hermida, J. R., ... Urbánek, T. (2012). Testing practices in the 21 st century: Developments and
European psychologists’ opinions. European Psychologist, 17(4), 300–319. http://dx.doi.org/10.1027/1016-9040/a000102.
Fischer-Klein, B. (2007). Evaluation eines Therapieprogramms zur Behandlung von Rechenschwäche. [Evaluation of a therapy program to treating learning disabilities in math.].
(Unpublished Dissertation). Available from EBSCOhost pdx databaseFaculty of Psychology, University of Vienna.
Flanagan, D. P., Fiorello, C. A., & Ortiz, S. O. (2010). Enhancing practice through application of Cattell–Horn–Carroll theory and research: A third method approach to
specific learning disability identification. Psychology in the Schools, 47(7), 739–760. http://dx.doi.org/10.1002/pits.20501.
Fletcher, J. M., & Vaughn, S. (2009). Response to intervention: Preventing and remediating academic difficulties. Child Development Perspectives, 3(1), 30–37. http://dx.
doi.org/10.1111/j.1750-8606.2008.00076.x.
Francis, D. J., Fletcher, J. M., Stuebing, K. K., Lyon, G. R., Shaywitz, B. A., & Shaywitz, S. E. (2005). Psychometric approaches to the identification of LD. Journal of
Learning Disabilities, 38(2), 98–108. http://dx.doi.org/10.1177/00222194050380020101.
Fuchs, L. S., & Fuchs, D. (2002). Mathematical problem-solving profiles of students with mathematics disabilities with and without comorbid reading disabilities.
Journal of Learning Disabilities, 35(6), 564–574. http://dx.doi.org/10.1177/00222194020350060701.
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(3), 561–576. http://dx.doi.org/10.1037/
a0014701.
Fuchs, L. S., Fuchs, D., Craddock, C., Hollenbeck, K. N., Hamlett, C. L., & Schatschneider, C. (2008). Effects of small-group tutoring with and without validated
classroom instruction on at-risk students’ math problem solving: Are two tiers of prevention better than one? Journal of Educational Psychology, 100(3), 491–509.
http://dx.doi.org/10.1037/0022-0663.100.3.491.
Fuchs, L. S., Fuchs, D., Stuebing, K., Fletcher, J. M., Hamlett, C. L., & Lambert, W. (2008). Problem solving and computational skill: Are they shared or distinct aspects
of mathematical cognition? Journal of Educational Psychology, 100(1), 30–47. http://dx.doi.org/10.1037/0022-0663.100.1.30.
Geary, D. C., Hamson, C. O., & Hoard, M. K. (2000). Numerical and arithmetical cognition: A longitudinal study of process and concept deficits in children with
learning disability. Journal of Experimental Child Psychology, 77(3), 236–263. http://dx.doi.org/10.1006/jecp.2000.2561.
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(1), 206–223. http://dx.doi.org/10.1037/a0025398.
Gelman, R., & Gallistel, C. R. (2004). Language and the origin of numerical concepts. Science, 306(5695), 441–443. http://dx.doi.org/10.1126/science.1105144.
Giofrè, D., & Cornoldi, C. (2015). The structure of intelligence in children with specific learning disabilities is different as compared to typically development children.
Intelligence, 52, 36–43. http://dx.doi.org/10.1016/j.intell.2015.07.002.
Hale, J., Alfonso, V., Berninger, V., Bracken, B., Christo, C., Clark, E., ... Yalof, J. (2010). Critical issues in response-to-intervention, comprehensive evaluation, and
specific learning disabilities identification and intervention: An expert white paper consensus. Learning Disability Quarterly, 33(3), 223–236. http://dx.doi.org/10.
1177/073194871003300310.
Jiménez Gonzalez, J. E., & Espinsel, A. I. G. (1999). Is IQ-achievement discrepancy relevant in the definition of arithmetic learning disabilities? Learning Disability
Quarterly, 22(4), 291–301. http://dx.doi.org/10.2307/1511263.
Jordan, N. C., & Hanich, L. B. (2003). Characteristics of children with moderate mathematics deficiencies: A longitudinal perspective. Learning Disabilities
Research & Practice, 18(4), 213–221. http://dx.doi.org/10.1111/1540-5826.00076.
Jordan, N. C., Hanich, L. B., & Kaplan, D. (2003). A longitudinal study of mathematical competencies in children with specific mathematics difficulties versus children
with comorbid mathematics and reading difficulties. Child Development, 74(3), 834–850.
Kaufmann, L., Handl, P., & Thöny, B. (2003). Evaluation of a numeracy intervention program focusing on basic numerical knowledge and conceptual knowledge: A
pilot study. Journal of Learning Disabilities, 36(6), 564–573. http://dx.doi.org/10.1177/00222194030360060701.
Kaufmann, L., Mazzocco, M. M., Dowker, A., von Aster, M., Göbel, S. M., Grabner, R. H., ... Nuerk, H.-C. (2013). Dyscalculia from a developmental and differential
perspective. Frontiers in Psychology, 4. http://dx.doi.org/10.3389/fpsyg.2013.00516.

65
K. Lambert, B. Spinath Research in Developmental Disabilities 72 (2018) 56–66

Klassen, R. M., Neufeld, P., & Munro, F. (2005). When IQ is irelevant to the definition of learning disabilities: Australian school psychologists? Beliefs and practice.
School Psychology International, 26(3), 297–316. http://dx.doi.org/10.1177/0143034305055975.
Klauer, K. J., & Phye, G. (2008). Inductive reasoning: A training approach. Review of Educational Research, 78, 85–123. http://dx.doi.org/10.3102/
0034654307313402.
Krajewski, K., & Schneider, W. (2009). Early development of quantity to number-word linkage as a precursor of mathematical school achievement and mathematical
difficulties: Findings from a four-year longitudinal study. Learning and Instruction, 19(6), 513–526. http://dx.doi.org/10.1016/j.learninstruc.2008.10.002.
Kroesbergen, E. H., & Van Luit, J. E. (2003). Mathematics interventions for children with special educational needs. Remedial & Special Education, 24(2), 97–114.
http://dx.doi.org/10.1177/07419325030240020501.
Lambert, K., & Spinath, B. (2013). Veränderungen psychischer Belastung durch die Förderung von rechenschwachen Kindern und Jugendlichen. [Changes in psy-
chological strain via support for children and adolescents with deficits in mathematics]. Zeitschrift für Kinder-und Jugendpsychiatrie und Psychotherapie, 41(1),
23–34. http://dx.doi.org/10.1024/1422-4917/a000207.
Lambert, K., & Spinath, B. (2014). Do we need a special intervention program for children with mathematical learning disabilities or is private tutoring sufficient?
Journal of Educational Research Online, 6(1), 68–93.
Lambert, K., & Spinath, B. (2017). Conservation abilities, visuospatial skills and numerosity processing speed: Association to math achievement and math difficulties in
elementary school children. Journal of Learning Disabilities, 1–3. http://dx.doi.org/10.1177/0022219417690354 online first.
Lipsius, M., & Petermann, F. (2009). Vergleich zwischen HAWIK-III und HAWIK-IV für Testanwender. [Comparison of HAWIK-III and HAWIK-IV for test users].
Diagnostica, 55(4), 255–261. http://dx.doi.org/10.1026/0012-1924.55.4.255.
Maehler, C., & Schuchardt, K. (2011). Working memory in children with learning disabilities: Rethinking the criterion of discrepancy. International Journal of Disability
Development and Education, 58(1), 5–17. http://dx.doi.org/10.1080/1034912X.2011.547335.
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. http://dx.doi.org/10.1007/s11881-003-0011-7.
McGill, R. J., Styck, K. M., Palomares, R. S., & Hass, M. R. (2015). Critical issues in specific learning disability identification: What we need to know about the PSW
model. Learning Disability Quarterly, 39(3), 159–170. http://dx.doi.org/10.1177/0731948715618504.
Melchers, P., Preuß, U., & Kaufman, A. S. (2007). Kaufman assessment battery for children (K-ABC). Leiden: PITS.
Mercer, C. D., & Miller, S. P. (1992). Teaching students with learning problems in math to acquire, understand, and apply basic math facts. Remedial & Special
Education, 13(3), 19–35. http://dx.doi.org/10.1177/074193259201300303 61.
Miciak, J., Taylor, P., Denton, C. A., & Fletcher, J. M. (2015). The effect of achievement test selection on identification of learning disabilities within a patterns of
strengths and weaknesses framework. School Psychology Quarterly, 30(3), 321–334. http://dx.doi.org/10.1037/spq0000091.
Moeller, K., Fischer, U., Cress, U., & Nuerk, H.-C. (2012). Diagnostics and intervention in developmental dyscalculia: Current issues and novel perspectives. In Z.
Breznitz, O. Rubinsten, V. J. Molfese, D. L. Molfese, Z. Breznitz, O. Rubinsten, V. J. Molfese, & D. L. Molfese (Eds.). Reading, writing, mathematics and the developing
brain: Listening to many voices (pp. 233–275). New York, NY, US: Springer Science+Business Media.
Murphy, M. M., Mazzocco, M. M. M., Hanich, L. B., & Early, M. C. (2007). Cognitive characteristics of children with mathematics learning disability (MLD) vary as a
function of the cutoff criterion used to define MLD. Journal of Learning Disabilities, 40(5), 458–478. http://dx.doi.org/10.1177/00222194070400050901.
Passolunghi, M. C., & Cornoldi, C. (2008). Working memory failures in children with arithmetical difficulties. Child Neuropsychology, 14(5), 387–400. http://dx.doi.
org/10.1080/09297040701566662.
Petermann, F., & Petermann, U. (Eds.). (2007). Hamburg-Wechsler-Intelligenztest für Kinder – IV (HAWIK-IV) [Hamburg-Wechsler intelligence test for children – IV.]. Bern:
Huber.
Pimperton, H., & Nation, K. (2010). Understanding words, understanding numbers: An exploration of the mathematical profiles of poor comprehenders. British Journal
of Educational Psychology, 80(2), 255–268. http://dx.doi.org/10.1348/000709909X477251.
Poletti, M. (2016). WISC-IV intellectual profiles in Italian children with specific learning disorder and related impairments in reading, written expression, and
mathematics. Journal of Learning Disabilities, 49(3), 320–335. http://dx.doi.org/10.1177/0022219414555416.
Raven, J. C., Court, J. H., & Raven, J. (1984). Manual for Raven’s progressive matrices and vocabulary scales. London: H. K. Lewis.
Reuhkala, M. (2001). Mathematical skills in ninth-graders: Relationship with visuo-spatial abilities and working memory. Educational Psychology, 21(4), 387–399.
http://dx.doi.org/10.1080/01443410120090786.
Rousselle, L., Palmers, E., & Noël, M.-P. (2004). Magnitude comparison in preschoolers: What counts? Influence of perceptual variables. Journal of Experimental Child
Psychology, 87(1), 57–84.
Schleifer, P., & Landerl, K. (2011). Subitizing and counting in typical and atypical development. Developmental Science, 14(2), 280–291. http://dx.doi.org/10.1111/j.
1467-7687.2010.00976.x.
Schlotmann, A. (2004). Warum Kinder an Mathe scheitern, wie man Rechenschwäche wirklich heilt. [Why children fail at maths, how to really cure mathematical learning
disabilities.], Vol. 2. Hirschberg: Supperverlag.
Sonntag, W. (2004). Der Einfluss des Klauerschen Denktrainings auf mathematisches Denken und Lernen von lernbehinderten Sonderschülern. [The impact of the
Cognitive Training for Youths on the mathematical thinking and learning of students with special educational needs. Zeitschrift für Pädagogische Psychologie, 18(2),
101–111. http://dx.doi.org/10.1024/1010-0652.18.2.101.
Stanovich, K. E. (2005). The future of a mistake: Will discrepancy measurement continue to make the learning disabilities field a pseudoscience? Learning Disability
Quarterly, 28(2), 103–105.
Stuebing, K. K., Fletcher, J. M., LeDoux, J. M., Lyon, G. R., Shaywitz, S. E., & Shaywitz, B. A. (2002). Validity of IQ-discrepancy classifications of reading disabilities: A
meta-analysis. American Educational Research Journal, 39(2), 469–518. http://dx.doi.org/10.3102/00028312039002469.
Swanson, H. L., & Sachse-Lee, C. (2001). Mathematical problem solving and working memory in children with learning disabilities: Both executive and phonological
processes are important. Journal of Experimental Child Psychology, 79(3), 294–321. http://dx.doi.org/10.1006/jecp.2000.2587.
Swanson, H. L., Lussier, C. M., & Orosco, M. J. (2015). Cognitive strategies, working memory, and growth in word problem solving in children with math difficulties.
Journal of Learning Disabilities, 48(4), 339–358. http://dx.doi.org/10.1177/0022219413498771.
Taylor, W. P., Miciak, J., Fletcher, J. M., & Francis, D. J. (2017). Cognitive discrepancy models for specific learning disabilities identification: Simulations of psy-
chometric limitations. Psychological Assessment, 29(4), 446–457. http://dx.doi.org/10.1037/pas0000356.
Tewes, U., Rossmann, P., & Schallberger, U. (Eds.). (2002). Hamburg-Wechsler-Intelligenztest für Kinder – dritte Auflage (HAWIK-III). [Hamburg-Wechsler intelligence test
for children](3rd ed.). Bern: Huber.
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(3), 240–256. http://dx.doi.org/10.1177/0022219414538520.
Van Luit, J. E. H., & Naglieri, J. A. (1999). Effectiveness of the MASTER program for teaching special children multiplication and division. Journal of Learning
Disabilities, 32(2), 98–107. http://dx.doi.org/10.1177/002221949903200201.
von Aster, M. G., & Shalev, R. S. (2007). Number development and developmental dyscalculia. Developmental Medicine and Child Neurology, 49(11), 868–873.
Watkins, M. W. (2000). Cognitive profile analysis: A shared professional myth. School Psychology Quarterly, 15(4), 465–479. http://dx.doi.org/10.1037/h0088802.
Wechsler, D. (1991). The Wechsler intelligence scale for children (3rd ed.). San Antonio, TX: The Psychological Corporation.
Wechsler, D. (2003). (4th ed.). Wechsler intelligence scale for childrenVol. WISC-IV. San Antonio, TX: The Psychological Corporation.
White, W. A. (1988). A meta-analysis of the effects of direct instruction in special education. Education & Treatment of Children, 11(4), 364–374.
Wittwer, J. (2008). Warum wirkt Nachhilfe? Hinweise aus der Forschung zum Einzelunterricht. [Why does tutoring work? Indications from the research on one-on-one
teaching.]. Zeitschrift für Pädagogik, 54(3), 416–432.
World Health Organisation (2005). ICD: Classification of mental and behavioural disorders: Clinical decriptions and diagnostic guidelines. Geneva, Switzerland: World
Health Organisatio.
Zirkel, P. A., & Thomas, L. B. (2010). State laws and guidelines for implementing RTI. Teaching Exceptional Children, 43(1), 60–73.

66