Κριτικές του άρθρου

Writing, reading and graphic-motor deficits in Greek children with developmental dyslexia

Andreas N. Bessas - Μπέσσας Ανδρέας

Abstract

The dyslexia studies examine the process of writing in children in order to explore the causal factors leading to dyslexia, the clarification of symptoms occurring in children and the attendant disorders that affect their course in school. The study results proclaim the existence of a close relationship between the symptoms of dyslexia and the abilities of processing input, processing of language, senses, perception, attention, memory, as well as the development of speech in pre-school years.

The present study aims at locating the errors in written texts that constitute the diagnostic criteria for the classification of dyslexia in 3 categories. The categorization of groups leads to the location of the respective types of dyslexia.

The investigation uses the following methods: dictation of text, copy of text and free writing of text by description of a subject picture. To identify the types of errors in writing are generally studied 137 children with dyslexia from 2^nd and 3^rd grade who attend regular schools. For implementation of the experiment are used texts from the school textbooks for the 2^nd and 3^rd grade. Identification criteria are the errors of children admitted in writing.

The categorization of errors in groups leads to the location of the respective types of dyslexia. The errors in text writing can be categorized in groups, according to the disorder in the acoustic-cognitive, visual-spatial or linguistic process of information. These errors are associated with children’s gender and reading abilities; although, the errors in text writing seem to relate to problems in text reading, bad handwriting and the ability to use lineal limits. These four abilities (writing, reading, handwriting and lineal skills) constitute relevant elements of the written language pathology that appears in children with developmental dyslexia.

Περίληψη

Οι μελέτες σχετικά με την δυσλεξία ερευνούν την διαδικασία της γραφής σε παιδιά προκειμένου να διερευνηθούν οι αιτιακοί παράγοντες που οδηγούν στην δυσλεξία, την αποσαφήνιση των συμπτωμάτων που εμφανίζονται στα παιδιά και τις συνοδές διαταραχές που επηρεάζουν την πορεία τους στο σχολείο. Τα αποτελέσματα των ερευνών καταδεικνύουν την ύπαρξη στενής σχέσης μεταξύ των συμπτωμάτων της δυσλεξίας και των ικανοτήτων επεξεργασίας των εισερχόμενων πληροφοριών, γλωσσικής επεξεργασίας, αισθήσεων, αντίληψης, μνήμης, καθώς και την εξέλιξη του προφορικού λόγου στην παιδική ηλικία.

Η παρούσα μελέτη στοχεύει στον εντοπισμό των λαθών κατά την γραφή κειμένου που αποτελούν διαγνωστικά κριτήρια για την ταξινόμηση της δυσλεξίας σε 3 κατηγορίες. Η κατηγοριοποίηση σε ομάδες οδηγεί στον εντοπισμό των αντίστοιχων τύπων δυσλεξίας.

Για την έρευνα έχουν χρησιμοποιηθεί οι εξής μέθοδοι: υπαγόρευση κειμένου, αντιγραφή κειμένου και περιγραφή σύνθετης εικόνας. Για την ταυτοποίηση των λαθών γραφής εξετάστηκαν 137 παιδιά με δυσλεξία 2^ης και 3^ης τάξης που παρακολουθούν κανονικό σχολείο. Για την εκτέλεση της έρευνας χρησιμοποιήθηκαν κείμενα από τα σχολικά βιβλία της 2^ης και 3^ης τάξης. Τα κριτήρια πιστοποίησης των λαθών είναι τα λάθη που παρουσιάζονται στην γραφή.

Η κατηγοριοποίηση των λαθών σε ομάδες οδηγεί στον εντοπισμό των αντίστοιχων τύπων δυσλεξίας. τα λάθη στην γραφή κειμένου μπορούν να κατηγοριοποιηθούν ανάλογα με την διαταραχή στην ακουστική-γνωστική, οπτικο-χωρική και γλωσσική επεξεργασία των πληροφοριών. Τα λάθη αυτά σχετίζονται με το φύλο των παιδιών και τις αναγνωστικές τους ικανότητες. Όμως, τα λάθη κατά την γραφή κειμένου φαίνεται ότι σχετίζονται με τα προβλήματα ανάγνωσης, τον γραφικό χαρακτήρα και τις γραφοκινητικές δεξιότητες των παιδιών. Αυτές οι τέσσερεις ικανότητες αποτελούν σημαντικούς παράγοντες για την παθολογία του γραπτού λόγου στα παιδιά με εξελικτική δυσλεξία.

Introduction

The dyslexia has been an object of study for the global scientific community for more than a century. Current studies examine the process of writing in children in order to explore the causal factors leading to dyslexia, the clarification of symptoms occurring in children and the attendant disorders that affect the course of children in school. The study results proclaim the existence of a close relationship between the symptoms of developmental dyslexia and the abilities of processing input, processing of language, senses, perception, attention, memory, as well as the development of speech in pre-school years.

According to Boder (1), “it may be said that dyslexic children are found to be consistently more dysorthographic than dyslexic; the three reading-spelling patterns reveal that most of the errors made by dyslexic children do not occur at random, but in patterns of errors; identification of the three reading-spelling patterns helps to clarify why the classic dyslexic errors do not distinguish subtypes among dyslexic children and even tend to obscure their heterogeneity; all of the classic errors, notably the static and kinetic reversals, have been observed in all three of our subtypes; it would appear, therefore, that although the classic errors are valuable diagnostic signs, especially significant in older children, they are not invariable concomitants of developmental dyslexia; aside from the classic reversals, most of the dyslexic errors fall into two main groups-the intelligible phonetic errors and the unintelligible dysphonetic errors” (1).

According to, “it may be said that dyslexic children are found to be consistently more dysorthographic than dyslexic; the three reading-spelling patterns reveal that most of the errors made by dyslexic children do not occur at random, but in patterns of errors; identification of the three reading-spelling patterns helps to clarify why the classic dyslexic errors do not distinguish subtypes among dyslexic children and even tend to obscure their heterogeneity; all of the classic errors, notably the static and kinetic reversals, have been observed in all three of our subtypes; it would appear, therefore, that although the classic errors are valuable diagnostic signs, especially significant in older children, they are not invariable concomitants of developmental dyslexia; aside from the classic reversals, most of the dyslexic errors fall into two main groups-the intelligible phonetic errors and the unintelligible dysphonetic errors”.

Based on current studies on developmental dyslexia, the errors of children’s writing and reading a text are classified in three categories:1) acoustic errors (errors due to deficit in acoustic-cognitive processing of information), 2) visual errors (errors due to deficit in visual-spatial processing of information) and 3) linguistic errors (errors in linguistic processing of information). A review of studies helps us understand the findings lead to the conclusion of a complex etiology and pathology not only in children but in adults with dyslexia as well.

As Ziegler illustrates (2) (2006) “it has been commonly agreed that developmental dyslexia in different languages has a common biological origin: a dysfunction of left posterior temporal brain regions dealing with phonological processes; it seems quite clear that children in all languages need to develop appropriate symbol–sound mapping for learning to read; in alphabetic languages, children learn to map letters onto phonemes; in logographic languages, however, children might need to learn the mapping of complex graphic-motor programs onto whole-word phonology; to perform these mappings, the brain uses different neural circuits depending on the grain size and transparency with which writing systems represent phonology; at a cognitive level, however, impaired phonology is still the major cause of dyslexia”. Gunnell & Parlow (3)associate “poor interhemispheric communication with everyday memory problems and poor phonological processing with poor working memory”. Caravolas, Hulme, & Snowling (4) accept that “skilled spelling requires a foundation in phonological transcoding ability which in turn enables the formation of orthographic representations”. Other theories illustrate that “poor comprehenders despite having adequate phonological decoding skills, have problems reading words that are typically read with support from semantics” (5). A study by Blomert & Mitterer (6) is compatible with the phonological-coding deficit theory. According to this theory “the deficit is in the phonetic transform from analogue neural response patterns to digital lexical/phonological representations”. According to Lavidor, Johnston, & Snowling (7) “explaining the between-hemisphere differences in the relative engagement and specialization of the different representations, enhanced sensitivity to orthographic cues is developed in some cases of dyslexia when a normal, phonology-based left hemisphere word recognition processing is not achieved”. Ziegler, Perry, Ma-Wyatt, Ladner, & Schulte-Korne (8) claim that “dyslexics exhibited a reading speed deficit, a specific non-word reading deficit, and a phonological decoding mechanism that operates extremely slowly and serially”. Also Tremblay, Monetta, & Joanette (9) “complexity processing modulates the hemispheric dynamic associated with phonology and semantics differently”.

Results from investigation of Zaidel & Petters (10) “help clarify an outstanding debate about whether phonological recoding is necessary for reading either words or sentences; right hemispheres reading of single words without phonological recoding is neurologically possible and the absence of phonological recoding may reflect constraints in short-term verbal memories and, in turn, restrict them to reading phrases shorter than three words long; together, this suggests that phonetic recoding is necessary for sentence reading”.

Research of Beaton (11) indicates that “some dyslexic children and adults have problems in subtle aspects of speech perception or discrimination which might underpin their phonological problems”. According to Alain, et al. (12) (2005) “deficits in parsing acoustic events may contribute to speech perception problems often observed in older adults and in individuals with dyslexia proposed that the explicit registration and discrimination of acoustic objects require computations subsequent to the initial processing of sounds in the ascending pathway and primary acoustic cortex, and suggest that these computations might be carried out in the planum temporal; according to this model, neural activity in the planum temporale should vary with listeners’ likelihood of hearing one or two acoustic objects”. Boets, Wouters, van Wieringen, & Ghesquiere (13) argue that “there is most evidence to situate the core of the reading and spelling problem at the level of higher-order phonological processing; low-level acoustic and speech perception problems are relatively over-represented in the group of literacy-impaired subjects and they might possibly aggravate the phonological and literacy problem”. Study of Bogliotti, Serniclaes, Messaoud-Galusi, & Sprenger-Charolles (14) confirmed “the relationship between reading skills and speech perception; examination of individual performances showed that both the deficit in categorical perception and the concomitant increase in allophonic sensitivity were fairly prevalent among children affected by dyslexia”. Study of Brunellière, Dufour, Nguyen, & Frauenfelder (15) shows that “the regional variability in the speech input to which listeners are exposed affects the perception of speech sounds in their own accent”. Gerrits and de Bree (16) find “lower speech perception and production performance in children at familial risk of dyslexia; perceptual discrimination between speech sounds belonging to different phoneme categories is better than that between sounds falling within the same category; this property, known as “categorical perception”, is weaker in children affected by dyslexia; dyslexic children use an allophonic mode of speech perception that, although without straightforward consequences for oral communication, has obvious implications for the acquisition of alphabetic writing. According to Serniclaes, Van Heghe, Mousty, Carre, & Sprenger-Charolles (17) “allophonic perception specifically affects the mapping between graphemes and phonemes, contrary to other manifestations of dyslexia, and may be a core deficit”.

According to Snowling (18) “at about the same time as the first epidemiological studies were being conducted, cognitive psychologists began comparing groups of normal and dyslexic readers in a range of experimental paradigms; these researchers pursued the then popular idea that dyslexia was a perceptual deficit and studies investigated visual perception, visual memory, cross-modal transfer between visual and verbal codes and perceptual learning, and other skills in relation to reading ability”. Facoetti & Molteni (19) argues that “the deficit of the mechanism subserving selection of stimuli (spatial attention) in the right visual field might determine some visual perceptual disorders that are frequently found in subjects with specific reading disorder or dyslexia”. Mycroft, Behrmann, & Kay (20) claims that “a longstanding and controversial issue concerns the underlying mechanisms that give rise to letter-by letter reading: while some researchers propose a prelexical, perceptual basis for the disorder, others postulate a postlexical, linguistic source for the problem; findings are consistent with significant visuoperceptual impairment in letter-by letter reading that adversely affects reading performance as well as performance on other non-reading tasks”. Lorusso, Facoetti, Pesenti, Cattaneo, Molteni, & Geiger (21) support that “the wide distribution of recognition, similar across the various subtypes of dyslexia, suggests a general characteristic of visual perception, and possibly a different visual-attentional mode”.

Snowling (18) accepts that “there is evidence that “dyslexic children have trouble with long-term verbal learning; this problem may account for many classroom difficulties, including problems memorizing the days of the week or the months of the year, mastering multiplication tables, and learning a foreign language; in a similar vein, this problem may be responsible for the poor vocabulary development often observed in dyslexic children”. Barber & Kutas (22) support that “during visual word recognition, the brain extracts various types of information presumed to be characterizing word representations; to that end, the visual (or word) processing system could segment words into a variety of sublexical units, associated with different information types (phonological, syllabic, morphological)”. Boder (1) has proposed that “one can distinguish dysphonetic dyslexics, whose deficiencies lie essentially in deciphering skills, dyseidetic ones who are mainly deficient in visual recognition and mixed cases combining the two deficiencies” (23).

Beaton (11) claims that “there is certainly a genetic connection between language lateralization and handedness although the exact nature of the relationship is controversial and handedness is probably related in some way to reading and dyslexia”.

Nicolson, Fawcett, & Dean (24) support that “disorders of cerebellar development can in fact cause the impairments in reading and writing characteristics of dyslexia, a view consistent with the recently appreciated role of the cerebellum in language-related skills”.

According to Beaton (11) “testosterone has been linked to the size of another brain structure which some have seen as having a role to play in dyslexic symptomatology; this structure is the corpus callosum”. Barber & Kutas (22) argues that “reading is not a genetically transmitted cognitive ability; it seems to involve the progressive specialization of specific brain areas; it has thus been possible to associate activity in different brain regions during reading with initial perceptual analysis, orthographic decoding, and phonological processing”. Study of Corbetta (25) “elucidates the pathophysiology of clinical brain disorders, that involve attentional and visual perceptual deficits, e.g. unilateral neglect, attentional deficit disorders, or dyslexia”. According to Mano, Osmon, & Klein (26) “the morphological factor could be interpreted as a construct representing perceptual decomposition of the word stem from prefix and suffix morphemes; their results add to the growing literature supporting the double-deficit hypothesis of dyslexia and have clinical implications for assessment of the orthographic aspects of dyslexia”. Baillieux, Vandervliet, Manto, Parizel, De Deyn, & Mariën (27) propose “a new hypothesis regarding the pathophysiological mechanisms of developmental dyslexia; given the sites of activation in the cerebellum in the dyslexic group, a defect of the intra-cerebellar distribution of activity is suspected, suggesting a disorder of the processing or transfer of information within the cerebellar cortex. According to Schulz, et al. (28) “developmental dyslexia is a specific disorder of reading acquisition characterized by a phonological core deficit; sentence reading is also impaired in dyslexic readers, but whether semantic processing deficits contribute is unclear; sentence reading was characterized by activation in a left-lateralized language network; semantic processing was characterized by activation in left-hemispheric regions of the inferior frontal and superior temporal cortex”. Results from a study of Steinbrink et al. (29), “based on a combined analysis of white and gray matter abnormalities, provide exceedingly strong evidence for a disconnection syndrome or dysfunction of cortical areas relevant for reading and spelling; this imbalance of neuronal communication between the respective brain areas might be the crucial point for the development of dyslexia”. According to Rae, et al. (30) “the relationship of cerebellar asymmetry to phonological decoding ability and handedness, together with previous finding of altered metabolite ratios in the cerebellum of dyslexics, suggest that there are alterations in the neurological organization of the cerebellum which relate to phonological decoding skills, in addition to motor skills and handedness”.

Experiment

The present study aims at locating the errors in written texts that constitute the diagnostic criteria for the classification of dyslexia in 3 different categories: acoustic, visual and linguistic dyslexia. The categorization of errors in groups of acoustic, visual and linguistic errors leads to the location of the respective types of dyslexia, which are analyzed according to the gender and grade of the children and their specific characteristics such as reading abilities and handwriting.

Method

The investigation uses the following methods: dictation of text, copy of text and free writing of text by description of a subject picture.

Subjects

To identify the types of errors in writing are generally studied 137 children with dyslexia from 2^nd and 3^rd grade who attend regular schools, 77 of which are boys and 60 girls, 72 children in 2^nd and 65 in 3^rd grade. 2^nd grade has 43 boys and 29 girls and 3^rd grade 34 boys and 31 girls.

Materials

For implementation of the experiment are used texts from the school textbooks for the 2^nd and 3^rd grade. Periods of the experiment were chosen so that the texts have already been taught to the children.

Procedure

The first text is given to the child for dictation with the instruction to write the text that will be dictated by the examiner. Examiner dictates every time no more than 3-4 words. Second text is given to the child in printed form with the instruction to copy it. For the description a black and white picture is selected, which depicts a playground and the child is given the instruction to write what he/she sees.

The time of the study was set at 15 minutes for each text, but some of the children needed more time for the copy from dictation or description of the picture, so the examination time differs from child to child.

Identification criteria

Identification criteria are the errors of children admitted in writing. Written texts are checked and the errors that have been described in each table are labeled in a certain way. Errors are grouped according to their individual characteristics, based on sound, visual or linguistic deficiency etiology.

Data analysis

In this particular experiment, therefore, that the data follow a regular distribution for the analysis used the distribution of χ² (non parametric statistics). Errors of children are classified each into 2, 3 or more categories according to a property (Robert & James, 1960). For implementation of this analysis the number of errors that children made for different types of text are collected and analyzed by gender, class and age of children during the study. For gender analysis a random number of 54 boys and girls is used. For education level analysis a random number of 60 children in 2^nd and 60 children in 3^rd grade is used. For reading ability level analysis a random number of 50 children with reading problems and 50 children without reading problems is used. For handwriting level analysis a random number of 50 children with bad handwriting and 50 children with good handwriting is used. For linear restrictions’ skills level analysis a random number of 49 children with deficits in linear restrictions’ skills and 49 children without deficits in linear restrictions’ skills is used. Children who have no data in the three texts are not included in the analysis.

For further investigation about the strength of the associations between means of the children’s errors in writing and children’s gender, grade, age, reading problems and handwriting SPSS one-way ANOVA statistics are used.

Results

The experiment includes data from the texts of 137 children in 2^nd and 3^rd grade. From total 77 boys and 60 girls, 72 boys and 55 girls have valid cases for acoustic errors (AE), 42 boys and 25 girls have valid cases for visual errors (VE) and 76 boys and 57 girls have valid cases for linguistic errors (LE). For AE the boys’ mean is 10.28 and the girls’ 7.13; for VE the boys’ mean is 2.76 and the girls’ 2.76; for LE the boys’ mean is 13.63 and the girls’ 9.07. From total 72 2^nd grade children and 65 3^rd grade children, 66 2-grade and 61 3-grade have valid cases for AE, 39 2-grade and 28 3-grade have valid cases for VE and 69 2-grade and 64 3-grade have valid cases for LE. For AE the 2-grade mean is 9.42 and the 3-grade 8.36; for VE the 2-grade mean is 3.21 and the 3-grade 2.14; for LE the 2-grade mean is 10.86 and the 3-grade 12.56. From total 57 children with reading problems (RP) and 80 without RP, 56 children with RP and 71 without have valid cases for AE, 31 children with RP and 36 children without have valid cases for VE and 57 children with RP and 76 children without have valid cases for LE. For AE the RP mean is 12.41 and the RP-not 6.15; for VE the RP mean is 3.32 and the RP-not 2.28; for LE the RP mean is 15.68 and RP-not 8.67. From total 96 children with bad handwriting (HW) and 41 children with good HW, 94 children with bad HW and 33 with good HW have valid cases for AE, 54 children with bad HW and 13 with good HW have valid cases for VE and 95 children with bad HW and 38 with good HW have valid cases for LE. For AE the bad HW mean is 10.29 and the good HW 5.00; for VE the bad HW mean is 2.59 and the good HW 3.46; for LE the bad HW mean is 13.49 and the good HW 7.13. From total 82 children with deficits in linear restrictions’ skills (LRS) and 55 without LRS, 80 children with LRS and 47 without have valid cases for AE, 46 children with LRS and 21 children without have valid cases for VE and 81 children with LRS and 52 without have valid cases for LE. For AE the LRS mean is 10.86 and the LRS-not 5.60; for VE the LRS mean is 2.63 and the LRS-not 3.05; for LE the LRS mean is 14.25 and the LRS-not 7.67.

Errors in writing of text

From total 137 children, 127 have valid cases for AE, 67 have valid cases for VE and 133 have valid cases for LE. AE are acoustic changes of letters, omission of letters, syllables or words, metathesis and antimetathesis of letters and syllables, errors in letter writing [j] and addition of letters, syllables or words. VE are visual changes of letters, repetition of letters, syllables or words, mirroring of letters, syllables or letter sequences, omission and repetitions of sequences. LE are errors in syntax and language usage, union and disunion of words, errors in use of punctuation marks and capital letters. Others errors (OE) in writing of text are mixed acoustic-visual changes of letters, grammatical-orthographical errors and errors in use of stresses.

95 children have valid cases for acoustic changes of letters (M=4.25), 104 children have valid cases for omission of letters, syllables or words (M=4.84), 37 children have valid cases for metathesis and antimetathesis of letters and syllables (M=1.84), 56 children have valid cases for addition of letters, syllables or words (M=2.00) and 34 children have valid cases for errors in letter writing [j] (M=1.32). 42 children have valid cases for visual changes of letters (M=2.21), 20 children have valid cases for repetition of letters, syllables or words (M=1.50), 29 children have valid cases for mirroring of letters, syllables or letter sequences (M=1.79), 9 children have valid cases for omission and repetitions of sequences (M=1.11). 90 children have valid cases for errors in syntax and language usage (M=2.60), 88 children have valid cases for union and disunion of words (M=6.40), 94 children have valid cases for errors in use of punctuation marks (M=4.32) and 79 children have valid cases for errors in use of capital letters (M=3.49). 41 children have valid cases for mixed acoustic-visual changes of letters (M=1.83), 131 children have valid cases for grammatical-orthographical errors (M=25.20) and 123 children have valid cases for errors in use of stresses (M=26.26).

Analysis of different types of errors in relation with gender and education level

In order to explore the relationship between the sum of the different types of errors in writing and children’s gender a distribution of χ² was conducted. Statistical analysis of acoustic, visual and linguistic errors in the writing of text depending on the gender of the children shows that there are statistically significant differences (χ²=6.65, df=2, N=54, p=.038), so that gender affects the types of errors in writing of text, and specifically the acoustic errors, with the boys to do most of them. Differences in boys and girls are significant and the boys have approximately 2 times more errors when writing a text than the girls. Visual and acoustic errors of boys have the same proportion. In writing of text the boys are expected to make many more errors. The acoustic errors have biggest frequency, which is almost as visual and linguistic errors together. However, this is not depending of the gender, because the sum of visual and linguistic errors in the girls was close to the number of acoustic errors, something which appears in boys too.

In order to explore the relationship between the sum of the different types of errors in writing and children’s grade a distribution of χ² was conducted. Statistical analysis of acoustic, visual and linguistic errors in the writing of text depending on the education level of children shows that there are no statistically significant differences (χ²=1.80, df=2, N=60, p=.405), so that the education level does not affect the types of errors in writing of text by the children. Errors of children at 2^nd and 3^rd grade do not have significant differences. Children while writing a text are expected to make the same mistakes as children in 2^nd and 3^rd grade. The sum of errors changes very little during increase of the level of education.

Table 1

Analysis of different types of errors in relation with children’s reading ability, handwriting skills and deficits in compliance with the linear constraints

In order to explore the relationship between the sum of the different types of errors in writing and children’s reading ability a distribution of χ² was conducted. Statistical analysis of acoustic, visual and linguistic errors in the writing of the readable text, depending on the reading skills shows that there are statistically significant differences (χ²=11.60, df=2, N=50, p=.003), so that the reading ability of children affects the types of errors when writing a text, and specifically the acoustic errors, with the children who have problems with reading doing most of them. Children who have trouble in reading made more errors when writing text, in the three categories. Most errors that children make are acoustic and visual have the lowest sum. This means that in children who have trouble in reading, the deficit prevails in acoustic processing of linguistic information, which leads to the emergence of acoustic errors when writing text. The relationship between writing and reading shows the difficulty of separating these processes, one affects the other. Ability of reading text determines the success of the children in writing.

In order to explore the relationship between the sum of the different types of errors in writing and children’s handwriting skills a distribution of χ² was conducted. Statistical analysis of acoustic, visual and linguistic errors in the writing of the text depending on the handwriting of children shows that there are statistically significant differences (χ²=10.58, df=2, N=50, p=.005), so that the handwriting of children affects the types of errors when writing text, and specifically linguistic errors, with the children who have ugly handwriting to make most of them. The children’s skills of writing letters associated with the occurrence of different types of errors when writing text. Children who have ugly handwriting made more errors in writing. These skills are abilities developed from pre-aged in drawing and construction tasks in nursery school. When children start school they learn the forms of letters and how to "draw" them. The child who does not write the correct form of letters meets major difficulties in the differentiation and storing letters, and other mental abilities that are associated with the proper knowledge of letters. On the other hand, teachers often say that children, who have problems with reading and writing, knowingly write a letter to another to mask their inability to write properly. However, it cannot be confirmed by the data of this experiment.

In order to explore the relationship between the sum of the different types of errors in writing and children’s linear restrictions’ skills (LRS) a distribution of χ² was conducted. Statistical analysis of acoustic, visual and linguistic errors in the writing of the text depending on the linear restrictions’ skills shows that there are statistically significant differences (χ²=15.18, df=2, N=49, p<.001), so that the appearance of distortion of words and lines affect the types of errors when writing text, and specifically acoustic errors, with children who do not comply with the restrictions make most of them. Poor ability to enforce restrictions on the line in the workbook of children led to an increase in the number of acoustic errors. These children have expressed in writing in the pathology of children who know how to write properly on the line.

Table 2

Analysis of different types of errors in writing among children’s gender, grade and age groups

In order to explore the means of the different types of errors in writing among the gender groups a one-way ANOVA was used. Children’s gender includes 2 grouping variables: boys and girls. Comparing the children’s gender groups on AE, VE and LE, no statistically significant difference was found among the AE (F(1,102)=3.211,p=.076), VE (F(1,51)=2.404,p=.127) and LE (F(1,108)=.959,p=.330). However, because the variances are unequal for VE, we are uncertain whether to trust results for VE. The boys’ mean for AE is 7.84, for VE 2.45 and LE 11.15. The girls’ means for AE is 5.70, VE 1.70 and LE 9.42. This indicates that the boys have a strongest pathology than the girls, but the differences are not big enough to understand the strength of those differences.

In order to explore the means of the different types of errors in writing among the grade groups a one-way ANOVA was used. Children’s grade includes 2 grouping variables: 2^nd grade and 3^rd grade. Comparing the children’s grade groups on AE, VE and LE, no statistically significant difference was found among the AE (F(1,102)=.032,p=.858), VE (F(1,51)=2.633,p=.111) and LE (F(1,108)=.680,p=.411). However, because the variances are unequal for VE, we are uncertain whether to trust results for VE. Children in 2^nd grade and children in 3^rd grade have approximately equal means for AE, 2^nd grade’s mean is 6.76 and 3^rd grade’s mean is 6.98. The means for VE are 2.52 for 2^nd grade and 1.75 for 3^rd grade. The means for LE are 9.64 for 2^nd grade and 11.09 for 3^rd grade. This indicates that more children in 2^nd grade make VE and more children in 3^rd grade make LE, but the differences are not big enough to understand the strength of those differences.

In order to explore the means of the different types of errors in writing among the age groups a one-way ANOVA was used. Children’s age includes 3 grouping variables: 7, 8 and 9 years old children. Comparing the children’s age groups on AE, VE and LE, no statistically significant difference was found among the AE (F(2,101)=1.144,p=.323), VE (F(2,50)=.370,p=.693) and LE (F(2,107)=.290,p=.749). The means for 7 years old children for AE is 5.94, for VE 2.12 and for LE 9.97. The means for 8 years old for AE is 6.78, for VE 2.36 and for LE 10.11. The means for 9 years old for AE is 8.52, for VE 1.83 and for LE 11.73. This indicates that the bigger children means are bigger than the smaller children, but the differences are not big enough to understand the strength of those differences.

In order to explore the differences between boys and girls in both grades and the means of the different types of errors in writing among the gender (by grade) groups a one-way ANOVA was used. Children’s gender (by grade) includes 4 grouping variables: boys in 2^nd grade, boys in 3^rd grade, girls in 2^nd grade and girls in 3^rd grade. Comparing the children’s gender (by grade) groups on AE, VE and LE, no statistically significant difference was found among the AE (F(3,100)=1.145,p=.335), VE (F(3,49)=1.671,p=.185) and LE (F(3,106)=.637,p=.593). However, because the variances are unequal for VE, we are uncertain whether to trust results for VE. The means of boys in 2^nd grade for AE is 7.45, for VE 2.84 and for LE 10.35. The means of boys in 3^rd grade for AE is 8.31, for VE 1.93 and for LE 12.11. The means of girls in 2^nd grade for AE is 5.70, for VE 1.90 and for LE 8.37. The means of girls in 3^rd grade for AE is 5.70, for VE 1.50 and for LE 10.10. This indicates that boy have mean values higher than girls for AE, VE and LE. AE and LE have higher mean values in 3^rd grade than 2^nd grade, but VE have higher mean values in 2^nd grade; although the differences are not big enough to understand the strength of those differences.

In order to explore the differences between little and big children in both grades and the means of the different types of errors in writing among the grade (by age) groups a one-way ANOVA was used. Children’s grade (by age) includes 4 grouping variables: little children in 2^nd grade, big children in 2^nd grade, little children in 3^rd grade and big children in 3^rd grade. Comparing the children’s grade (by age) groups on AE, VE and LE, no statistically significant difference was found among the AE (F(3,100)=1.278,p=.286), VE (F(3,49)=1.518,p=.222) and LE (F(3,106)=.314,p=.816). However, because the variances are unequal for VE, we are uncertain whether to trust results for VE. The mean for AE is 5.94 for little 2^nd grade children, 8.16 for big 2^nd grade children, 5.97 for little 3^rd grade children and 8.52 for big 3^rd grade children. The mean for VE is 2.12 for little 2^nd grade children, 3.00 for big 2^nd grade children, 1.67 for little 3^rd grade children and 1.83 for big 3^rd grade children. The mean for LE is 9.97 for little 2^nd grade children, 9.14 for big 2^nd grade children, 10.69 for little 3^rd grade children and 11.73 for big 3^rd grade children. This indicates that little children in both grades have approximately equal means for AE; for VE and LE the means for 2^nd grade and 3^rd grade are approximately equal; although, the differences are not big enough to understand the strength of those differences.

Analysis of different types of errors in writing among children’s RP groups

In order to explore the means of the different types of errors in writing among the reading problems (RP) groups a one-way ANOVA was used. Children’s RP includes 2 grouping variables: children with RP and children without RP. Comparing the children’s RP groups on AE, VE and LE, statistically significant difference was found among the AE (F(1,102)=4.383,p=.039), VE (F(1,51)=7.657,p=.008) and LE (F(1,108)=11.85,p=.001). However, because the variances are unequal, we are uncertain whether to trust these results. The mean of AE is 8.35 for children with RP and 5.84 for children without RP. The mean of VE is 2.91 for children with RP and 1.65 for children without RP. The mean of LE is 13.91 for children with RP and 8.05 for children without RP. This indicates that children with RP have bigger mean values than children without RP.

Table 3

In order to explore the differences between major and minor RP and the means of the different types of errors in writing among the RP (major or minor) groups a one-way ANOVA was used. Children’s RP (major or minor) includes 3 grouping variables: children with major RP, children with minor RP and children without RP. Comparing the children’s RP (major or minor) groups on AE, VE and LE, no statistically significant difference was found among the AE (F(2,101)=2.550,p=.083), but statistically significant difference was found among the VE (F(2,50)=4.347,p=.018) and LE (F(2,107)=5.839,p=.004). However, because the variances are unequal, we are uncertain whether to trust results. The means of children with major RP for AE is 8.62, for VE 2.29 and for LE 15.19. The means of children with minor RP for AE is 8.32, for VE 3.12 and for LE 12.93. The means of children without RP for AE is 5.73, for VE 1.63 and for LE 8.08. This indicates that children with RP have mean values higher than children without RP for AE, VE and LE. Children with major RP and children with minor RP have approximately equal mean values in AE; children with major RP have higher mean values than children with minor RP in LE; children with minor RP have higher mean values than children with major RP for VE. Games-Howell test indicates that there are no significant differences among groups except the differences among children with major RP and children without RP; in AE between children with major RP and children with minor RP MD=.304 (p=.986); between children with major RP and children without RP MD=2.892 (p=.192); between children with minor RP and children without RP MD=2.588 (p=.152); in VE between children with major RP and children with minor RP MD=-.839 (p=.526); between children with major RP and children without RP MD=.652 (p=.373); between children with minor RP and children without RP MD=1.492 (p=.091); in LE between children with major RP and children with minor RP MD=2.256 (p=.765); between children with major RP and children without RP MD=7.111 (p=.033); between children with minor RP and children without RP MD=4.854 (p=.098).

Table 4

Analysis of different types of errors in writing among children’s HW groups

In order to explore the means of the different types of errors in writing among the handwriting (HW) groups a one-way ANOVA was used. Children’s handwriting includes 2 grouping variables: children with dab HW and children with good HW. Comparing the children’s HW groups on AE, VE and LE, statistically significant difference was found among the LE (F(1,108)=4.790,p=.031), but not on AE (F(1,102)=3.583,p=.061) and VE (F(1,51)=.562,p=.457). The mean AE is 7.57 for children with bad HW and 5.07 for children with good HW; the mean VE is 2.26 for children with bad HW and 1.82 for children with good HW; the mean LE is 11.62 for children with bad HW and 7.52 for children with good HW. This indicates that children with bad HW have bigger mean values than children without bad HW.

Table 5

Analysis of different types of errors in writing among children’s LRS groups

In order to explore the means of the different types of errors in writing among the LRS groups a one-way ANOVA was used. Children’s LRS includes 2 grouping variables: children with deficits in LRS and children without deficits in LRS. Comparing the children’s LRS groups on AE, VE and LE, statistically significant difference was found among the LE (F(1,108)=4.768,p=.031), but not on AE (F(1,102)=2.389,p=.125) and VE (F(1,51)=.116,p=.735). The mean AE is 7.62 for children with LRS and 5.73 for children with not LRS; the mean VE is 2.23 for children with good LRS and 2.06 for children with bad LRS; the mean LE is 11.95 for children with good LRS and 8.13 for children with bad LRS. This indicates that children with good LRS have bigger mean values than children with bad LRS.

Table 6

Discussion

Acoustic, visual and linguistic errors in the writing of text represent categories of errors that are used to determine the forms of dyslexia. It is these errors to be identified as dyslexic. Therefore, they are analyzed in relation to gender and education level.

According to the definition of dyslexia, it is not the result of the absence of stimuli, sensory dysfunction and insufficient training and environmental factors, however, they can co-exist. Therefore, the symptoms of dyslexia (in this case errors when writing a text), are not a result of these causes. The education level and the quality of the training process are causal factors and dyslexic errors in writing of text are not their results. So, errors during writing of text to be identified as dyslexic must demonstrate that they have connection with the gender of children and have no connection with the level of education.

Analysis of different types of errors in relation with gender shows that there are significant differences between the sum of acoustic, visual and linguistic errors. Although, an analysis of AE, VE and LE means in gender, grade and age groups shows that there are not significant differences between groups. The boys have greater mean values than the girls, which indicate a strongest pathology in writing. However, the differences are not big enough for such a conclusion. The same result is valid for the grade and age analysis. In grade groups we see that the means for AE and LE are too close, while the VE mean for 2^nd grade is a little higher than the 3^rd grade’s mean. Combining gender and grade we see that there is no significant differences in AE, VE and LE means. Combining the grade and age also there is no significant differences between means.

Exploring the means of AE, VE and LE among RP groups, we see that children with RP have greater mean values than the children without RP. The strongest difference appears in LE and VE. By analysis the means of AE, VE and LE among HW and LRS groups we see that there are significant differences for LE, but not for AE and VE. This indicates that only LE strongly depend on the RP, HW and LRS. By separating the RP children group in 2 groups of children with major RP and children with minor RP and combining them with no-RP children’s group we see that VE and LE have significant differences among groups. AE means for children with major RP and children with minor RP are approximately equal. Although, the differences seems be located at the major RP and no-RP groups.

In conclusion we say that:

The errors in text writing of children of pre-school years can be categorized in groups of acoustic, visual and linguistic errors, according to their specific characteristics, which lead us to the categorization of the errors in acoustic, visual and linguistic errors, according to the disorder in the acoustic-cognitive, visual-spatial or linguistic process of incoming information.

Acoustic errors in writing of text are acoustic changes of letters, omission of letters, syllables or words, metathesis and antimetathesis of letters and syllables, errors in letter writing [j] and addition of letters, syllables or words. Visual errors in writing of text are visual changes of letters, repetition of letters, syllables or words, mirroring of letters, syllables or letter sequences, omission and repetitions of sequences. Linguistic errors in writing of text are errors in syntax and language usage, union and disunion of words, errors in use of punctuation marks and capital letters.

The errors in text writing seem to relate to problems in text reading, bad handwriting and the ability to use lineal limits. These four abilities (writing, reading, handwriting and lineal skills) constitute relevant elements of the written language pathology that appears in children with developmental dyslexia.

Further research and analysis needs to be done in order to clarify the appearance of different errors in text writing according to the type of text before any categorization of the types of developmental dyslexia appearance. Further research is necessary in order to discover a possible relation of the errors in text writing with errors in text reading of children with developmental dyslexia. We would expect that the children present the same errors in both writing and reading, but this research cannot possibly provide such a conclusion.

References

1. Boder, Elena. Developmental Dyslexia: a Diagnostic Approach Based on Three Atypical Reading-spelling Patterns. Developmental Medicine and Child Neurology. 8 1973, Vol. 15, 5, pp. 663-687. http://www.scopus.com/scopus/inward/record.url?eid=2-s2.0-0015900561&partnerID=40.

2. Ziegler, Johannes C. Do differences in brain activation challenge universal theories of dyslexia? Brain and Language. 9 2006, Vol. 98, 3, pp. 341–343. http://www.sciencedirect.com/science/article/B6WC0-4GP7P1X-1/2/0a7033b273d15051af027b95595dcebb.

3. Gunnell, Jessica G. and Parlow, Shelley E. Interhemispheric communication, phonological processing, and memory in dyslexia: An investigation of relationships among three domains. TENNET: Theoretical and Experimental Neuropsychology - XVII Annual Meeting, June 21-23, 2007, Montreal, Canada. Brain and Cognition. 6 2008, Vol. 67, Supplement 1, pp. S11–S47. http://www.sciencedirect.com/science/article/B6WBY-4SYCGCG-1J/2/c7b0ea12024e4363c7c2f72a338f908b.

4. Caravolas, Markéta, Hulme, Charles and Snowling, Margaret J. The Foundations of Spelling Ability: Evidence from a 3-Year Longitudinal Study. Journal of Memory and Language. 11 2001, Vol. 45, 4, pp. 751–774. http://www.sciencedirect.com/science/article/B6WK4-457VF1T-D/2/2e539cd5d709c6b391e0e779512948fc.

5. Nation, Kate and Snowling, Margaret J. Semantic Processing and the Development of Word-Recognition Skills: Evidence from Children with Reading Comprehension Difficulties. Journal of memory and language. 7 1998, Vol. 39, 1, pp. 85–101. http://www.sciencedirect.com/science/article/B6WK4-45KV6HH-5/2/8cbc04e22a051e9fc051531c3835c858.

6. Blomert, Leo and Mitterer, Holger. The fragile nature of the speech-perception deficit in dyslexia: Natural vs. synthetic speech. Brain and Language. 4 2004, Vol. 89, 1, pp. 21–26. http://www.sciencedirect.com/science/article/B6WC0-49SFF4N-3/2/6444b0ebe927a573769f8427d1c5decf.

7. Lavidor, Michal, Johnston, Rhona and Snowling, Margaret J. When phonology fails: Orthographic neighbourhood effects in dyslexia. Brain and Language. 3 2006, Vol. 96, 3, pp. 318–329. http://www.sciencedirect.com/science/article/B6WC0-4GVGTC2-1/2/e37b6787b71e215c0e332b2461cb34ca.

8. Ziegler, Johannes C., et al. Developmental dyslexia in different languages: Language-specific or universal? J. Experimental Child Psychology. 11 2003, Vol. 86, 3, pp. 169–193. http://www.sciencedirect.com/science/article/B6WJ9-49S3C95-1/2/45a14664c3ca168b0105351f3165167a.

9. Tremblay, Tania, Monetta, Laura and Joanette, Yves. Complexity and hemispheric abilities: Evidence for a differential impact on semantics and phonology. Brain & Language. 2 2009, Vol. 108, 2, pp. 67–72. http://www.sciencedirect.com/science/article/B6WC0-4V4KC5P-1/2/e87a6827efe2188f72a9e53ea98e7ddb.

10. Zaidel, Eran and Petters, Ann M. Phonological Encoding and ldeographic Reading by the Disconnected Right Hemisphere: Two Case Studies. Brain and Language. 11 1981, Vol. 14, 2, pp. 205-234. http://www.sciencedirect.com/science/article/B6WC0-4D6RF4T-34/2/f266b15f35777c1a85bf1320adedbaa1.

11. Beaton, Alan A. The Neurobiology of Dyslexia. [book auth.] Martin Turner and John Rack. The Study of Dyslexia. New York : Kluwer Academic/Plenum Publishers, 2004, pp. 35-76.

12. Alain, Claude, et al. Left thalamo-cortical network implicated in successful speech separation and identification. NeuroImage. June 2005, Vol. 26, 2, pp. 592 – 599.

13. Boets, Bart, et al. Auditory processing, speech perception and phonological ability in pre-school children at high-risk for dyslexia: A longitudinal study of the auditory temporal processing theory. Neuropsychologia. 2007b, Vol. 45, 8, pp. 1608–1620. http://www.sciencedirect.com/science/article/B6T0D-4MV0MFF-3/2/51e9753742ca0d8d71470f3766c71b01.

14. Bogliotti, C., et al. Discrimination of speech sounds by children with dyslexia: Comparisons with chronological age and reading level controls. Journal of Experimental Child Psychology. 10 2008, Vol. 101, 2, pp. 137–155. http://www.sciencedirect.com/science/article/B6WJ9-4SFHM3T-1/2/19e6f99059d7b5885fda543c09c5a4ec.

15. Brunellière, Angèle, et al. Behavioral and electrophysiological evidence for the impact of regional variation on phoneme perception. Cognition. 6 2009, Vol. 111, 3, pp. 390-396. http://www.sciencedirect.com/science/article/B6T24-4VYW6CW-1/2/50c64c08f0f9fa8a5606b5046f43c61c.

16. Gerrits, Ellen and de Bree, Elise. Early language development of children at familial risk of dyslexia: Speech perception and production. Journal of Communication Disorders. 5-6 2009, Vol. 42, 3, pp. 180-194. http://www.sciencedirect.com/science/article/B6T85-4TX33N5-2/2/d3e697145f96b986dbc0a883d92588db.

17. Serniclaes, Willy, et al. Allophonic mode of speech perception in dyslexia. Journal of Experimental Child Psychology. 4 2004, Vol. 87, 4, pp. 336-361. http://www.sciencedirect.com/science/article/B6WJ9-4C0SV2F-1/2/2e3ea13d0c54d96d7e1b1cd7b8e7503c.

18. Snowling, Margaret J. The Science of Dyslexia: A Review of Contemporary Approaches. [book auth.] Martin Turner and John Rack. The Study of Dyslexia. New York : Kluwer Academic/Plenum Publishers, 2004, pp. 77-90.

19. Facoetti, Andrea and Molteni, Massimo. The gradient of visual attention in developmental dyslexia. Neuropsychologia. 2001, Vol. 39, 4, pp. 352–357. http://www.sciencedirect.com/science/article/B6T0D-427JW9G-3/2/9fb1d063dd119f86522f3185aa7aa0b2.

20. Mycroft, Rachel H., Behrmann, Marlene and Kay, Janice. Visuoperceptual deficits in letter-by-letter reading? Neuropsychologia. 6 2009, Vol. 47, 7, pp. 1733-1744. http://www.sciencedirect.com/science/article/B6T0D-4VM446F-1/2/b8091c90b6dc1487722d4330a7d13fc4.

21. Lorusso, M. L., et al. Wider recognition in peripheral vision common to different subtypes of dyslexia. Vision Research. 9 2004, Vol. 44, 20, pp. 2413–2424. http://www.sciencedirect.com/science/article/B6T0W-4CN9KBH-1/2/54f66aaf5254c77c8272aa111061ec95.

22. Barber, Horacio A. and Kutas, Marta. Interplay between computational models and cognitive electrophysiology in visual word recognition. Brain Research Reviews. 1 2007, Vol. 53, 1, pp. 98-123. http://www.sciencedirect.com/science/article/B6SYS-4KMYKYH-1/2/370c0ffcd158fd0bbfb5f70abf6b1158.

23. Bartelson, Paul. The onset of literacy: Liminal remarks. Cognition. 11 1986, Vol. 24, 1-2, pp. 1-30. http://www.scopus.com/scopus/inward/record.url?eid=2-s2.0-0022814217&partnerID=40.

24. Nicolson, Roderick I., Fawcett, Angela J. and Dean, Paul. Developmental dyslexia: the cerebellar deficit hypothesis. Trends in Neurosciences. 9 1, 2001, Vol. 24, 9, pp. 508-511. http://www.sciencedirect.com/science/article/B6T0V-43PFXTH-M/2/8bb271d6b6166d28eb21d4a6ede1524f.

25. Corbetta, Maurizio M. FMRI studies of visual motion and attention. Academic Radiology. December 2001, Vol. 8, 12.

26. Mano, Q. R., Osmon, D. C. and Klein, L. Factor analysis of visuoperceptual-orthographic processing: Speed, Early Perception, Late–Post Perception, and Morphological Awareness. National Academy of Neuropsychology, Abstracts from the 25th Annual Meeting, Tampa, Florida, October 19–22, 2005. 2005, Vol. 20, pp. 805–950.

27. Baillieux, Hanne, et al. Developmental dyslexia and widespread activation across the cerebellar hemispheres. Brain & Language. 2 2009, Vol. 108, 2, pp. 122–132. http://www.sciencedirect.com/science/article/B6WC0-4TVFY8V-1/2/a8a72a7d9d3afa13ea2e01e0f4abca29.

28. Schulz, Enrico, et al. Impaired semantic processing during sentence reading in children with dyslexia: Combined fMRI and ERP evidence. NeuroImage. 5 15, 2008, Vol. 41, 1, pp. 153–168. http://www.sciencedirect.com/science/article/B6WNP-4S1C8C4-4/2/4b28de16b9052de70c8ed0823e3c9021.

29. Steinbrink, C., et al. The contribution of white and gray matter differences to developmental dyslexia: Insights from DTI and VBM at 3.0 T. Neuropsychologia. 11 2008, Vol. 46, 13, pp. 3170-3178. http://www.sciencedirect.com/science/article/B6T0D-4T2M664-2/2/29586e03ad83d01f90eb058ed828ead8.

30. Rae, Caroline, et al. Cerebellar morphology in developmental dyslexia. Neuropsychologia. 2002, Vol. 40, 8, pp. 1285-1292. http://www.sciencedirect.com/science/article/B6T0D-44KVSNH-1/2/bec8e5db3bfadc304e27c82f4f1d008f.