In Table 1, data regarding intelligence assessment of the three groups, performed through the WISC-III, can be verified. The gross scores and ratings obtained are shown. Specifically, through the analysis of gross scores, significant differences were observed among the three groups in RDI and PSI. On the other hand, when the groups were compared two by two, it was verified that the NCG showed higher scores than the DD, statistically significant, only in the PSI. When compared to developmental dyslexia, differences were observed in PSI, VIQ, CR and RDI, with significantly higher NCG scores. Statistically significant differences between DD and developmental dyslexia were not observed.

Table 1

WISC-III	Groups								Mann-Whitney
	DD			DYSLEXIA			NCG
	Gross Scores
	M	SD		M	SD		M	SD
VIQ	83.87	17.43		76.84	14.01		94.08	19.80	NCG > DYSLEXIA*
PIQ	146.62	33.00		135.76	32.87		164.75	37.25	-
TIQ	230.50	48.21		212.61	41.25		252.83	52.18	-
CR	60.75	15.07		54.84	11.91		69.08	18.72	NCG > DYSLEXIA*
POI	93.37	23.69		84.38	28.47		92.25	33.42	-
RDI a,*	23.12	3.56		22.61	3.15		26.83	4.15	NCG > DYSLEXIA*
PSI a,*	53.25	10.22		51.38	11.98		67.41	14.58	NCG > DD*
									NCG > DYSLEXIA**
	DD Classification - f (%)
	VS	SP		HA	AV		LA	BL	EL
VIQ	0 (0)	0 (0)		2 (25)	5 (62)		1 (12)	0 (0)	0 (0)
PIQ	0 (0)	0 (0)		4 (50)	4 (50)		0 (0)	1 (7)	0 (0)
TIQ	0 (0)	1 (12)		2 (25)	5 (62)		0 (0)	0 (0)	0 (0)
CR	0 (0)	0 (0)		2 (25)	6 (75)		0 (0)	0 (0)	0 (0)
POI	0 (0)	0 (0)		3 (37)	5 (62)		0 (0)	0 (0)	0 (0)
RDI	0 (0)	1 (12)		0 (0)	3 (37)		3 (37)	1 (12)	0 (0)
PSI	0 (0)	1 (12)		1 (12)	6 (75)		0 (0)	0 (0)	0 (0)
	Developmental Dyslexia Classification - f (%)
	VS	SP		HA	AV		LA	BL	EL
VIQ	1 (7)	0 (0)		10 (76)	0 (0)		2 (15)	0 (0)	0 (0)
PIQ	0 (0)	1 (7)		5 (38)	4 (30)		2 (15)	1 (7)	0 (0)
TIQ	0 (0)	1 (7)		3 (23)	9 (69)		0 (0)	0 (0)	0 (0)
CR	0 (0)	1 (7)		1 (7)	10 (76)		1 (7)	0 (0)	0 (0)
POI	0 (0)	2 (15)		4 (30)	4 (30)		3 (23)	0 (0)	0 (0)
RDI	0 (0)	0 (0)		1 (7)	9 (69)		2 (15)	1 (7)	0 (0)
PSI	0 (0)	1 (7)		2 (15)	8 (61)		2 (15)	0 (0)	0 (0)
	NCG Classification - f (%)
	VS	SP		HA	AV		LA	BL	EL
VIQ	4 (33)	2 (16)		1 (8)	2 (16)		3 (25)	0 (0)	0 (0)
PIQ	3 (25)	2 (16)		4 (33)	1 (8)		2 (16)	0 (0)	0 (0)
TIQ	4 (33)	1 (8)		4 (33)	2 (16)		1 (8)	0 (0)	0 (0)
CR	4 (33)	2 (16)		2 (16)	2 (16)		2 (16)	0 (0)	0 (0)
POI	2 (16)	2 (16)		2 (16)	4 (33)		2 (16)	0 (0)	0 (0)
RDI	3 (25)	2 (16)		2 (16)	3 (25)		1 (8)	1 (8)	0 (0)
PSI	3 (25)	3 (25)		2 (16)	4 (33)		0 (0)	0 (0)	0 (0)

Abbreviations: M: Mean; SD: Standard Deviation; f: Frequency; %: percentage; VS: Very Superior; SP: Superior; HA: High Average; AV: Average; LA: Low Average; BL: Borderline; EL: Extremely Low; ^a: Kruskal-Wallis; * : significant difference at the p ≤ 0,05 level; ** : significant difference at the p ≤ 0,01 level.

Regarding the instruments selected for attention assessment, when comparing the gross scores obtained by the three groups in both subtests (Coding and Symbol Search), significant differences were found between them. When comparing groups two by two, NCG scores were significantly higher in relation to DD, in both subtests. In the same way, this result occurred in relation to the developmental dyslexia. However, when comparing the DD and developmental dyslexia groups, again no significant differences were found between them in both subtests (Table 2).

Table 2

Instruments	Groups								Mann-Whitney
	DD			DYSLEXIA			NCG
	Gross Scores
	M	SD		M	SD		M	SD
Attention
Coding a,*	34.12	6.87		32.61	7.98		43.00	9.95	NCG > DD *
									NCG > DYSLEXIA**
Symbol Search a,*	19.12	3.72		18.76	4.58		24.41	6.25	NCG > DD *
									NCG > DYSLEXIA*
Executive Functions
Digit Span (backward order)	3.28	1.38		3.84	2.51		4.37	2.44	-
Arithmetic	13.12	1.80		13.38	2.06		14.33	1.96	-
PRBC	37.62	1.76		35.69	3.94		37.00	5.16	NCG > DYSLEXIA*
Stroop Test
Time (sec) - Card 1 a,*	25.25	4.83		30.07	7.17		22.75	4.61	NCG > DYSLEXIA**
No. errors	0.25	0.46		0.15	0.55		0.00	0.00	-
Time (sec) - Card 2	35.50	7.23		33.69	9.47		32.91	9.55	-
No. errors	0.50	1.06		0.07	0.27		0.75	1.42	-
Time (sec) - Card 3	41.75	13.24		45.46	31.37		40.66	14.92	-
No. errors	0.87	1.12		0.23	0.43		0.83	1.40	-
WCST
No. categories completed	4.37	1.99		4.15	1.81		5.08	1.31	-
No. tests (max. 128)	114.12	17.85		120.69	17.84		115.50	17.44	-
Persistent Errors (max 128)	29.00	24.61		35.00	22.82		22.33	13.81	-
Total correct responses	71.12	15.99		62.23	16.58		73.33	14.81	-
Total errors	43.00	25.19		58.46	25.41		42.25	21.83	-
Failure to maintaining the context	1.00	1.41		0.46	0.87		1.58	1.24	NCG > DYSLEXIA*
Memory
Digit a,**	3.28	1.38		3.84	2.51		4.37	2.44	NCG > DD **
									NCG > DYSLEXIA**
RCFT
Copy	35.00	29.27		27.30	26.81		34.16	28.02	-
Memory	26.25	17.87		25.53	24.45		35.00	36.36	-
Time copy	44.62	22.89		35.15	28.18		37.91	20.27	-
Time memory	35.87	23.55		22.25	11.02		31.08	19.89	-

Abbreviations: M: Mean; SD: Standard Deviation; ^a: Kruskal-Wallis; * : significant difference at the p ≤ 0,05 level; ** : significant difference at the p ≤ 0,01 level.

In Table 2 it is also possible to observe the EF data. When comparing the gross scores obtained in the Digit span subtest (backward order) among the three groups, no significant differences were verified among them, and this also happens when the groups were compared two by two. In the Arithmetic subtest, this same result is observed. In the PRBC test, no significant differences were found among the three groups. However, when comparing groups two by two, there was a significant difference between NCG and developmental dyslexia, with higher NCG scores. In the Stroop test, significant differences were observed among the three groups only in time needed to name the first card colors. In the comparison of groups two by two, only one difference between groups was found, which was between NCG and developmental dyslexia and in the comparison of this same variable of the test, with shorter times obtained by NCG, that is, superior performances of this group.

Regarding the WCST test, when comparing the scores obtained by the three groups, no significant differences were observed in any of the measures evaluated. When comparing groups two by two, significant differences were identified only in the category of ‘failure to maintaining the context’ between NCG and developmental dyslexia, with higher NCG performances. Thus, there were no significant differences between the DD and developmental dyslexia groups in the instruments used for EF assessment.

Still in Table 2, the data related to memory assessment are observed. In the Digit Span subtest (forward order) there was a significant difference among the three groups when compared at once. In the comparison of groups two by two, statistically significant higher NCG scores were verified when compared to DD and developmental dyslexia. No difference between DD and developmental dyslexia. When comparing the three groups in the RCF test, no significant differences were found among them, it was the same case when groups two by two were compared.

Finally, Table 3 shows the data obtained in the instruments selected for the evaluation of educational skills. It was initially verified in the APT that when comparing the three groups there were statistically significant differences among them in the Reading and Writing subtests and in the total of that test. When comparing groups two by two, it was verified that there was no significant difference between NCG and DD in any of the subtests, nor in the total. On the other hand, there was significant difference between the NCG and developmental dyslexia groups in Reading, Writing and Total, with the first showing higher scores. Between the groups with DD and developmental dyslexia, the same pattern of differences was observed, with DD showing higher scores. In the Arithmetic Test there were no significant differences among the three groups. When comparing groups two by two, no significant differences were found between any of the groups.

Table 3

Instruments	Groups								Mann-Whitney
	DD			DYSLEXIA			NCG
	Gross Scores
	A	SD		A	SD		A	SD
APT
Reading a,***	63.25	6.79		44.76	21.17		66.66	2.22	NCG > DYSLEXIA***
									DD > DYSLEXIA**
Writing a,**	24.12	5.16		13.92	8.28		25.41	5.46	NCG > DYSLEXIA***
									DD > DYSLEXIA**
Arithmetic	13.37	5.18		11.23	4.95		13.66	3.65	-
Total a,**	100.87	14.20		69.92	30.71		105.75	8.94	NCG > DYSLEXIA***
									DD > DYSLEXIA**
Arithmetic Test	41.12	14.56		35.53	12.83		43,08	10,78	-

Abbreviations: A: Average; SD: Standard Deviation; ^a: Kruskal-Wallis; * : Significant difference at the p ≤ 0,05 level; ** : significant difference at the p ≤ 0,01 level; *** : significant difference at the p ≤ 0,001 level.

In the intelligence assessment, no differences were observed among the groups in the TIQ, which indicates a certain intragroup homogeneity. These results were expected, since children with Specific Learning Disorders do not have performance deficits in this variable (APA, 2013). The differences found among the groups in the PSI and RDI will be discussed in detail when attention, memory and executive functions are addressed, since the subtests that comprise them were part of the assessment of such abilities. The lower performance in the VIQ and CR found in the group with dyslexia, when compared to the non-case group, was in agreement with some studies that also pointed out such differences, probably due to the linguistic impairment, characteristic of the picture in question (Lima, Azoni, & Ciasca, 2013; Moura, Simões, & Pereira, 2014). Differences between the groups with dyslexia and dyscalculia were not observed and no study was found that indicated similar or discrepant results in these categories.

In the Attention assessment, there was inferior performance of both the groups with dyscalculia and dyslexia when compared to the non-case group. These findings meet national and international literature, which showed impairments in this ability, in these disorders, when compared to groups of individuals without school difficulties (Ashkenazi et al., 2009; Ziegler et al., 2010). Regarding dyslexia, studies have speculated that this would be due to deficits in the functioning of the parietal zones of the right hemisphere associated (Peyrin et al., 2012), or even the left hemisphere in the temporal zones, and thus, the theory of phonological processing would be even more supported (Landerl & Willburger, 2010). In dyscalculia there is a frequent observation that hypofunctioning in the right parietal lobe (zone related to numeric and quantity sense) would explain both the disorder and the attention deficit (Furman & Rubinsten, 2012; Rubinsten & Henik, 2009). No national and international studies have been found that investigated the comparison of these two disorders in this ability. These results indicated that there were no differences between such groups.

Regarding the EF, it was verified that the group with dyscalculia did not differ from the non-case group in any of the tests. The group with dyslexia differed from the non-case group in pseudoword repetition, in the color-naming time of the Stroop test and in the “failure to maintaining the context” of the WCST test. However, broader deficits were expected in both disorders, according to the literature, and in most of the subtests that were within the assessment protocol of this ability (as in Digit Span and Arithmetic) (Menghini et al., 2010). Researches indicate that the prefrontal cortex would be the most related to EF (Stuss, 2011). Hypothetically, it can be said that impairments in Specific Learning Disorders in such a brain zone would be expected, as well as in EF, since dysfunctions in posterior zones are verified (as shown in the previous paragraph). According to Luria (1981), for the proper functioning of frontal zones it is necessary that zones after it are intact, which is not the case in these cases. When the diagnostic groups were compared in the EF, no significant differences between them were again found. This result was compatible with the study by Landerl et al. (2009), which also indicated that in digit repetition the groups did not differentiate. On the other hand, Wang et al. (2012) pointed out differences in the Stroop test, in boards involving numbers and letters, which did not occur here in relation to the words (stimulus that composed the test structure used here).

In the memory assessment, there was lower performance of the group with dyscalculia and dyslexia in short-term auditory memory when compared to the non-case group. Studies indicate that this result is expected when these groups are compared to children without difficulties (Attout et al., 2015; Bogaerts et al., 2014). In short-term visual memory, no differences were found among any of the groups. This was not expected, according to the literature, especially when nosological groups were compared to children without learning difficulties (Bacon & Handley, 2014; Szucs, Devine, Soltesz, & Gabriel, 2013).

Theoretically, there are some postulates affirming that attention and executive abilities act collaterally and / or even are integral mechanisms of the functioning and maintenance of memory, both short and long term (Bull, Espy, & Wiebe, 2008). As both skills are at a disadvantage in Specific Learning Disorders, then, consequently in these groups, impairments in different types of memory must be found. This statement corroborates, to a certain extent, the data found here, and which pointed out lowered performances in attention and in some executive aspects. Studies aiming at comparing both disorders, which are discussed here and regarding memory, were not found, but it was found that both disorders did not differ statistically.

Unfortunately, studies aiming at comparing neuropsychological abilities of developmental dyslexia and dyscalculia were found to be scarce in the literature. In any case, what could explain the fact that they did not differ in any of the skills selected here to compose the neuropsychological profile (intelligence, attention, executive functions and memory)? Some hypotheses have been elaborated: (a) the functional impairments in posterior zones and characteristic of each disorder would alter the assessed abilities in the same way, in general, independently of the stimuli used in the instruments and even in the use of these skills in daily life (both linguistic and mathematical); (b) the disorders are explained by impairments in very close processing brain areas, and thus, they would show a similar neuropsychological profile in skills investigated here. These areas would involve left perisylvian zones, left occipito-temporal fusiform gyrus, fronto-parietal zones (Ramus et al., 2013, Rosenberg-Lee et al., 2015, Vicario, Rappo, Pepi, Pavan & Martino, 2012).

According to our findings, what differentiated both situations were topographically the school difficulties involved. Among them (writing, reading and arithmetic), there was a statistically significant lower performance in writing and reading of the group with dyslexia when compared to the group with dyscalculia, this also happened when the group with dyslexia was compared to the non-case group. Somehow, such results were expected, since dyslexia mainly involves written language impairments (APA, 2013; Ramus et al., 2013). So, when the tasks were involving specific reading and writing school skills, unlike when the neuropsychological abilities with linguistic stimuli were assessed, differences between the disorders were observed.

However, the school skill in arithmetic of the group with dyscalculia differing from the group with dyslexia and non-case group was also expected, with a statistically significant lower performance, but did not happen. According to the DSM-V (APA, 2013), a minimum of 1.5 standard deviation below the mean can be expected, but it is also justified that, although the use of psychometric instruments is performed to guide the diagnosis, any significant performance measure below the range may be considered arbitrary and clinical assessment should prevail. In this case, the means obtained by the group with dyscalculia were smaller than the group with dyslexia, but not enough to indicate statistical differences.

Interestingly, the group with dyslexia had lower mean scores in arithmetic than the group with dyscalculia. Studies in literature indicating the same results were not found. In this study, the children of the group with dyslexia did not have comorbidity with dyscalculia and showed inferior performance (although not statically significant) than those who only had this last situation. Thus, some hypotheses are raised for these results: (a) the greater impairment in neuropsychological abilities of executive functions of the group with dyslexia, when compared to the non-case group in relation to the one with dyscalculia, as seen here, would favor greater impairments in arithmetic (Geary et al., 2012). (b) the linguistic impairments of dyslexia would favor inferior performance in arithmetic skills, as research has also shown (Träff & Passolunghi, 2015), even lower than those seen in dyscalculia.

In any case, this research fulfilled its objective of comparing the neuropsychological and school profiles of children with dyscalculia and developmental dyslexia, in order to verify if the performances of both groups differed in more detail, in addition to the diagnostic criteria already established. The results showed that the groups did not differentiate as to the neuropsychological abilities investigated (intelligence, attention, executive functions and memory), but only in reading and writing school skills, with the group with dyslexia showing inferior performances. The hypothesis raised was thus confirmed, and only partially, in relation to these skills.

Reduced performances in neuropsychological abilities involving mathematical content (for example, in WISC-III Digit Span and Arithmetic subtests) in the group with dyscalculia compared to the group with dyslexia, and lower performances in neuropsychological abilities involving reading (for example, in the Pseudoword Repetition Test for Brazilian Children and Stroop Test) in the group with dyslexia when compared to the group with dyscalculia. It was also expected lower performance in the APT Arithmetic subtest and in the Arithmetic test in the group with dyscalculia in relation to the one with dyslexia.

These findings are important for health, education and research. For the first area, it could be observed that Specific Learning Disorders could have a generalized impact on the functioning of the investigated neuropsychological abilities (except in intelligence), demonstrating that they need strong interventions at the multi / interdisciplinary level.

For the education area, these results become relevant because they mainly demonstrate that the condition of dyscalculia may go unnoticed in the age groups and grades studied. According to the data found, the differences in the instruments that assessed the school skills in arithmetic were subtle among children without school difficulties and those with this condition. It should be emphasized, therefore, that attention to such subtleties in educational processes must be made for an early diagnosis. On the other hand, children with dyslexia would already be more easily identified (there were discrepant results in reading and writing skills when compared to the group without school difficulties).

As for the research in the field, the data point out to more careful analysis in evaluation processes. The first interesting point to note is regarding the historically used performance discrepancies between Intellectual Quotient (IQ) and performance in school tests in the performance of the diagnoses of specific learning disorders (APA, 2013). As seen here, there was no such difference mainly between the groups with dyscalculia and without school difficulties. Relying only on these data, without a qualitative analysis of clinical basis, should not be the only way for a consistent assessment. A second point of analysis to be surveyed is also suggested, from this point of analysis: criterion validity studies involving these conditions in neuropsychological tests and assessment of school skills. It was possible to identify that none of the instruments used in this research, which are mostly very frequently used in the Brazilian neuropsychological assessment, had these studies. Further research on these psychometric procedures is advised.

Therefore, the limitations found in this study are mainly related to the number of individuals investigated, which made it impossible to analyze the effect of the variables gender and grade. In future studies, it is expected that there will also be an increase in the number of instruments used to investigate the skills described here and that these have greater ecological validity. In addition, it is proposed that the neuropsychological abilities that compose the profile investigated here should be expanded, especially with linguistic skills (such as phonological processing) and in a longitudinal way (as already suggested by Silva, Moura, Wood & Haase, 2015), so that differences between dyscalculia and dyslexia are even better delineated.