Dyslexia and eye movement analysis
Eye movement analysis (OMA) is one of the most effective methods for studying reading. Compared with traditional behavioral techniques, eye movement analysis has many advantages. First, eye m over ent does not require an additional task, which avoids disturbing the normal reading process and enables a realistic measurement of the reader’s reading process. Second, eye-movement methods provide instantaneous measurements of the reader’s reading process and can simultaneously provide a number of temporally and spatially accurate indicators, such as eye position, gaze duration, and gaze sequence, that describe how the reader is processing the content at each moment. Initially, research on eye movements in reading focused on the basic parameters and characteristics of eye movements. With the development of cognitive psychology and eye-movement technology, eye-movement studies of reading have been intensified and the research field has been expanded. In addition to examining the basic characteristics of eye movements of different types of readers, researchers have begun to explore the relationship between eye movements and cognitive processing during reading.
Psychologists have argued that on an immediate basis, eye movements reflect a variety of cognitive processing activities and are also influenced by a variety of cognitive factors, such as the close relationship between eye movements and cognitive activities such as attention, anticipation, memory, and reasoning. Peng et al. compared the eye-movement trajectories of Chinese readers and English readers in reading English and Chinese stories and found that cognitive factors were more important than perceptual factors in determining readers’ eye-movement patterns and gaze retention. Unfortunately, although research on the relationship between cognitive processing and eye movement behavior in reading has been conducted for more than 100 years, and several theoretical models of eye movement control have been developed to explain it, there is no universally accepted theoretical model that can explain the relationship between all eye movement parameters and cognitive processing in reading because reading itself is a very complex cognitive process and it is not easy to represent this process through eye movements. Therefore, there is no universally accepted theoretical model to explain the relationship between all eye movement parameters and cognitive processing activities during reading.
As a special group of children with reading disorder RD who is significantly impaired in literacy and reading comprehension, it is important to understand the basic characteristics of their oculomotor parameters during reading and the possible associated cognitive processing factors to reveal the nature of their reading difficulties. From June 2004 to March 2006, a research group conducted an eye movement experiment with 19 Chinese-speaking RD children and 19 matched normal children, using short Chinese texts as targets. The results showed that children with Chinese RD had abnormal eye movement patterns in reading texts, mainly in terms of longer average gaze duration and smaller average eye hopping amplitude. Two of the eye movement indicators, mean gaze duration and mean eye-hopping amplitude, are significant predictors of the reading of texts by RD children. However, it is not clear why these abnormal eye movements occur in Chinese-speaking RD children, what these indicators reflect, and what cognitive factors influence them. Therefore, in order to further explore the possible cognitive factors associated with the abnormal eye movement patterns of Chinese-speaking RD children, and to provide a direction and basis for future research and intervention, we believe that further regression analyses are necessary to correlate the subjects’ cognitive processing with their eye movement parameters. Currently, the Wechsler Intelligence Test (WIT) is the most commonly used test to assess the cognitive abilities of RD children, and the verbal and operational subscales of the WIT reflect a range of different cognitive components. In this study, the Wechsler Intelligence Test scores of 19 Chinese RD children were correlated with their oculomotor index of text reading.
The RD group: According to the ICD10 diagnostic criteria, the developmental history, onset history, clinical observation, homework survey, parents’ and teachers’ evaluation, and intelligence assessment, 19 children with RD in grades 2-6 were diagnosed by the Child Developmental Behavior Center of the Third Affiliated Hospital of Sun Yat-sen University from June 2004 to March 2006, including 17 males and 2 females, aged 7 to 13. The children in the group also met the following criteria: The children also met the following criteria: normal learning experience, but with language learning difficulties and poor learning ability (below the 5th percentile in the class). The children also met the following criteria: normal learning experience, but poor learning ability (below the 5th percentile in the class). The IQ of the IQ test was >70. There was no obvious visual or auditory impairment on the @test. The same direct medical history and clinical observation were used to exclude emotional disorders, attention deficit hyperactivity disorder, and other neuropsychiatric disorders.
Control group: 19 normal children of the same period were selected for the 1:1 matching study based on age, grade, gender, parental culture, and family economic status. The Pupil Rating Scale Revised Screening for Learning Disabilities PRS, the C owners Teacher Scale and Parent Scale, the Chinese Wechsler Intelligence Scale for Children PRS, and the Chinese Wechsler Intelligence Scale for C hid ilities PRS were administered to both groups before the test. Intelligence Scale for Children and a parent questionnaire on children’s reading and writing skills were completed by parents.
The Eye link II High-Speed Eye-Tracker (SR research L td1, Canada), with a 21″ LCD display, was used to display the designed vocabulary of targets in sequence. The eye movements analyzed included: first gaze time, total gaze time, mean gaze time, gaze frequency, mean eye beat amplitude, eye beat frequency, eye beat distance, and gaze recall (in this study, gaze recall refers to the number of times the gaze point shifts between the two interest areas of the text and the question), a number of ignored words (the number of words in the text that are not attended to), and pupil diameter. The test begins with an average illumination level of 200100 lx, the subject’s eyes are 70 cm from the screen, the head is fixed horizontally and the system is practiced for 2-5 m in. Target display. The oculomotor starts recording immediately.
Eye movement test target material
Two short texts in Chinese characters, one from a narrative text for after-school reading in a Primary 2 textbook, with simple content and a topic sentence. The text is presented in 5 lines, with 8 sentences and 146 characters. The questions are presented on the screen at the same time as the text. There are 3 questions, each with 4 options (only one of which is correct), in 6 lines, with 79 characters. The whole essay has a perspective of approximately 11102 O ~ 26153 O. The Chinese characters in the essay are in Song font and each character has a perspective of approximately 1 O. Essay 2 is from a short essay in Neuropsychology on testing semi-lateral neglect. Five questions are also presented, asking for correct or incorrect judgments. The text is presented from 131,04O to 241,19O and the Chinese characters are in Song font, with each character having a perspective of approximately 1O. The subject is asked to read the text carefully and answer the questions.
Wechsler Intelligence Scale for Children China Revised
The scale includes 11 subtests: knowledge, classification, arithmetic, vocabulary, comprehension, number breadth, fill-in-the-blank, picture arrangement, wooden block pattern, piecing together, coding, and the scale scores for each subtest can be calculated separately according to age.
(V IQ), Operational IQ (PIQ) and Total IQ (FIQ).
Statistical methods
The oculomotor has its own eye movement analysis software, which automatically records the data. The results of the Wechsler IQ test were analysed using SPSS111 0 statistical software, mainly by means of paired t-tests and multiple linear regression analysis of the correlations between the two groups of children and the oculomotor index.
Results
Children with RD scored significantly lower on the Knowledge, Categorization, Vocabulary and Coding subtests than the control group, as well as on the Verbal IQ and Total IQ scores. The children’s verbal comprehension and attention/memory scores were significantly lower than those of the control group
The children had no significant deficits in perceptual organization.
| 量表 | RD 组 ( n = 19) | 对照组 ( n = 19 ) | t值 | P 值 |
| 知识 ( I) | 8 土 3 | 10 土3 | – 21 18 | 01040 |
| 分类 ( S) | 9 土 2 | 11 土3 | – 31 47 | 01003 |
| 算术 ( A ) | 9 土 3 | 11 土3 | – 11 72 | 01 100 |
| 词汇 ( V ) | 11 土 2 | 13 土3 | – 3 06 | 01007 |
| 领悟 ( C ) | 11 土 3 | 12 土2 | – 01 64 | 01 530 |
| 数字广度 ( D ) | 8 土 2 | 9 土3 | – 11 17 | 01 260 |
| 填图 ( PC ) | 10 土 2 | 10 土3 | – 01 54 | 01 600 |
| 图片排列 ( PA ) | 9 土 4 | 10 土4 | – 0 50 | 01 620 |
| 木块图案 ( BD ) | 10 土 3 | 11 土4 | – 11 09 | 01 290 |
| 图形拼凑 ( OA ) | 11 土 2 | 11 土3 | 01 27 | 01790 |
| 编码 ( CD ) | 8 土 3 | 11 土4 | – 21 30 | 01030 |
| 言语智商 ( V IQ ) | 98 土13 | 111 土14 | – 3 26 | 01004 |
| 操作智商 ( P IQ ) | 97 土16 | 104 土19 | – 11 42 | 01 170 |
| 全智商 ( F IQ ) | 97 土12 | 109 土16 | – 2 72 | 01010 |
| 言语理解因子 | 101 土14 | 111 土13 | – 2 52 | 0 020 |
| 知觉组织因子 | 99 土15 | 103 土17 | – 01 77 | 01450 |
| 注意 /记忆因子 | 89 土13 | 102 土17 | – 21 83 | 01010 |
| 变量 | B | S1E | Beta | t值 | P 值 |
| 词汇 | – 51 55 | 21 18 | – 01 32 | – 21 55 | 01010 |
| 图形拼凑 | 51 15 | 1178 | 01 32 | 21 89 | 01005 |
| 数字广度 | 41 94 | 2107 | 01 27 | 21 38 | 01020 |
| 知识 | – 31 73 | 11 85 | – 01 27 | – 21 02 | 01050 |
Based on the results of previous studies, two eye movement indicators, mean eye gaze duration and mean eye skip amplitude, were found to be significant predictors of article reading in children with RD. In order to further investigate which cognitive factors were associated with the mean eye gaze time and mean eye skip amplitude, multiple linear regression analyses were conducted using the mean eye gaze time and mean eye skip amplitude as the dependent variables Y and the 11-item C W ISC test as the independent variables for the two groups. Vocabulary and knowledge were negatively correlated with each other (Table 2). In the analysis of the effect on the mean eye beat amplitude, two sub-tests were included in the regression equation: block diagram and knowledge, where block diagram was negatively correlated with mean eye beat amplitude and knowledge were positively correlated with it.
IQ tests are an important psychodiagnostic technique that not only assesses a person’s level of intelligence, but also to some extent reflects other neuropsychopathological conditions associated with the patient, and are therefore widely used in neuropsychological practice. The most commonly used assessment of cognitive ability in RD is the Wechsler Intelligence Test. Studies have suggested that the verbal and operational scale subtests of the Wechsler Intelligence Test reflect a range of different cognitive components, and Mason noted a specific pattern in the W ISC CR subtests of RD children, in which each child has unbalanced scores on different cognitive functions. The results of this study found that RD children scored significantly lower on the Knowledge, Categorization, Vocabulary, and Coding subtests than the control group, and the differences were statistically significant. Among these four subtests, knowledge measured children’s knowledge and interest in the outside world, categorization measured abstract generalization, vocabulary measured the breadth of word knowledge, and coding measured short-term memory, attention, and hand-eye coordination. This indicates that Chinese RD children have poor abstract generalization ability, poor short-term memory, attention and hand-eye coordination, and a narrow range of knowledge and lexical knowledge. In addition, the verbal IQ, total IQ, verbal comprehension factor, and attention/recall factor scores of RD children were significantly lower than those of the control group, and the differences were statistically significant (P < 0.05). This indicates that children with RD have lower verbal IQ than normal children and have deficits in verbal comprehension and attention/memory, but no significant deficits in perceptual organization. This is similar to the results of our previous study. The results of the present study reaffirm M as son’s conclusion that RD children have unbalanced scores in different cognitive functions, mainly in verbal comprehension and attention/memory, but no significant deficits in perceptual organization. The results of this study showed that RD children’s scores of knowledge, categorization, and vocabulary in response to left-brain processing were significantly lower than those of the control group, and the difference was statistically significant. This may indicate that Chinese RD children have an imbalance between the two hemispheres, with left-brain processing being worse than right-brain processing.
In a previous study, two oculomotor indicators, mean eye gaze duration and mean eye skip amplitude during article reading, were found to be significant predictors for children with RD. In order to investigate which cognitive factors were related to the mean gaze duration and mean eyebounce amplitude, multiple linear regression analyses were conducted using the mean gaze duration and mean eyebounce amplitude as dependent variables Y and the 11 C W ISC subtests as independent variables for the two groups of children. The results showed that the mean gaze time was influenced by vocabulary (which measures the breadth of word knowledge), pictorial assembly (which measures the ability to use visual organization, coordination of visual movements, and the ability to perceive part-whole relationships), numerical breadth (which measures attention and short-term memory), and knowledge (which measures the child’s knowledge and interest in the outside world), with the scores of pictorial assembly and numerical breadth being positively correlated with each other. The scores of pictorial spelling and number breadth were positively correlated with each other, while vocabulary and knowledge were negatively correlated. This indicates that there are four main cognitive factors that influence the average gaze duration: visual cognitive processing ability (including visual organization, coordination of visual actions, and the relationship between perceptual parts and the whole); attention and short-term memory ability; verbal knowledge; and children’s knowledge. The stronger the child’s visual cognitive processing and short-term memory and attention skills, the narrower the child’s knowledge and lexical knowledge, and the longer the child’s average gaze time when reading an article. The results of the 11-item IQ test showed that children with RD had deficits in vocabulary and knowledge, but were normal in piecing together figures and number breadth. Therefore, this study suggests that the prolonged attention span of children with RD may be mainly related to their narrow knowledge and lexical knowledge. Research in reading psychology suggests that people need two types of information when reading, visual information and non-visual information. These two types of information interact with each other to produce reading comprehension. The acquisition of visual information is the external process of reading, while the systematic knowledge that people store in their brains is the internal process of reading. Psychologists refer to this in-memory contextual knowledge as a schema. It consists of information acquired during learning and experience, and is a structure of knowledge that exists in the person’s long-term memory and is not created at the moment of reading a written symbol. The comprehension of reading is supported by the existing knowledge system in the memory, and people receive and process the acquired linguistic information based on their own knowledge and experience, establish connections between the materials through reasoning, add the missing information, and finally achieve a reasonable interpretation of the linguistic material. Therefore, RD children’s narrow knowledge of words will lead to poor comprehension of words, and their ability to integrate existing knowledge with current linguistic information will be affected by their narrow knowledge. This is consistent with our hypothesis based on the oculomotor index. Some case studies have suggested that children with RD may have reading difficulties mainly at the word level, but not at the reading comprehension level. Based on the results of the present study, it is reasonable to believe that children in the RD group have both word-level processing and text integration deficits. The mean eye-hopping amplitude is the average distance between eye-hopping crossing points. During the eye bounce, spatial and temporal information of the stimulus can be obtained, but it is almost impossible to form a clear image of the stimulus, so the eye bounce can achieve a rapid search of the visual field and selection of stimulus information. In the published article, we speculated that Chinese RD children used a small amplitude of eye hopping during article reading, indicating that their attention span during article reading was narrow and they might use a local processing mode instead of global processing. The results of the multiple linear regression analysis showed that the mean eye-hopping amplitude was influenced by the block diagram (a test of visual-spatial analysis and synthesis) and the knowledge subtest, with the block diagram being negatively correlated with the mean eye-hopping amplitude and the knowledge being positively correlated with it. This indicates that the distance of eye bounce is influenced by the ability of visuospatial analysis and synthesis as well as the modulation of the subjects’ existing knowledge schema. The better the visuospatial analysis and synthesis ability and the narrower the knowledge, the smaller the amplitude of eye jump. Based on the results of the 11-item IQ test, RD children did not have significant deficits in their block diagrams; therefore, it was concluded that the low mean eye-hopping amplitude of RD children during article reading was mainly related to their narrow knowledge. It is hypothesized that the narrow knowledge of RD children directly affects their top-down internal processing of reading, such as anticipation, reasoning ability, and strategy, which may be inferior to those of normal children, thus causing RD children to pay attention to small localized eye jumps when reading. In summary, children with RD have unbalanced scores in different cognitive functions, mainly in verbal comprehension and attention/memory, but no significant deficits in perceptual organization. The abnormal eye movements during article reading were associated with a narrow knowledge base and word knowledge, which may reflect some deficits in top-down internal processing in this group of children.
Reference
[ 1] Li Xiuhong, Jing Jin, Zou Xiaobing, et al. An oculomotor trial study of article reading in Chinese dyslexic children [ J] Chinese Mental Health, 2007, 21 ( 6 ): 362 – 365
[ 2] Zhang Huifen: Visual perception, knowledge schema, and fast reading training [ J] Chinese Language Learning, 2000, ( 2 ): 42 – 48
[ 3] Peng DL, Orchard LN, S tern JN E valuation of eye movement variables of Chinese and American readers [ J]56- 60[ 4] Kato M, Miyash ita N, H ikosakaO, et an eye movements in monk eys w ith local spam in e depletion in the caudate nucleus I Deficits inspan taneous saccades [ J] J Neurosc,i 1995, 15 ( 1 Pt 2): 912 – 927
[ 5] World Health Organization ICD 10 Classification of Mental and Behavioural Disorders [M ] Beijing: People’s Health Press, 1993: 192- 193
[ 6] Wu Tan Zhang, Zhou Yong, L ing Wang, et al Difficulties in authority organization as a possible cau se o f reading back ward ness [ J] Nature, 1998, 271 ( 3): 746- 747 [ 7] Nature, 1998, 271 ( 3): 746- 747 : 746- 747
[ 7] Song Fang, Su Linyan, Zhu Yan Clinical application of Conners’ parent-teacher questionnaire in children aged 3-7 years [ J] Chinese Journal of Child Health, 2004, 12 ( 5 ): 376- 378
[ 8] Jing Jin, Hai Yan, Deng Guifen, et al. Revision and evaluation of the Learning Disability Screening Scale