FAMOUS CHILD QUOTE
A child is fed with milk and praise.
Mary Lamb
I would like to thank each of my class members for such a spirit feel discussion. Each of you kept the discussions fun and exciting. I truly enjoyed learning about the development of a child. I wish each of you much success as you go forward. Take care!
Audrey Winters
Tuesday, April 19, 2011
Saturday, April 9, 2011
SCHOOL AGE CHILDREN ASSESSED IN CHINA
ASSESSED CHILDREN IN CHINA
I believe a young child's development should be assessed. The reason being, the teacher has to have a starting point. This will help the teacher understand were the child is in terms of academics. Teachers refer to this as a base-line.
The additional ideas and suggestions I have to help children with problems be dysexia, try having the child tested earlier than their 5 birthday.
Journal of Child Psychology and Psychiatry 52:2 (2011), pp 204–211 doi:10.1111/j.1469-7610.2010.02299.x
Early predictors of dyslexia in Chinese children: familial history of dyslexia, language delay, and cognitive profiles
Catherine McBride-Chang,1 Fanny Lam,2 Catherine Lam,2 Becky Chan,2 Cathy Y.-C. Fong,3 Terry T.-Y. Wong,1 and Simpson W.-L. Wong4
1The Chinese University of Hong Kong, China; 2Child Assessment Service, Department of Health, Hong Kong, China; 3The University of Hong Kong, China; 4University of Oxford, UK
Background: This work tested the rates at which Chinese children with either language delay or familial history of dyslexia at age 5 manifested dyslexia at age 7, identified which cognitive skills at age 5 best distinguished children with and without dyslexia at age 7, and examined how these early abilities predicted subsequent literacy skills. Method: Forty-seven at-risk children (21 who were ini- tially language delayed and 26 with familial risk) and 47 control children matched on age, IQ, and mothers’ education were tested on syllable awareness, tone detection, rapid automatized naming, visual skill, morphological awareness, and word reading at age 5 and subsequently tested for dyslexia on a standard Hong Kong measure at age 7. Results: Of those with an early language delay, 62% subsequently manifested dyslexia; for those with familial risk, the rate of dyslexia was 50%. Those with dyslexia were best distinguished from those without dyslexia by the age-5 measures of morphological awareness, rapid automatized naming, and word reading itself; other measures did not distinguish the groups. In a combined regression analysis across all participants, morphological awareness uniquely explained word reading accuracy and rapid automatized naming uniquely explained timed word reading at age 7, with all other measures statistically controlled. Separate stepwise regression analyses by group indicated that visual skill uniquely explained subsequent literacy skills in the at-risk group only, whereas tone and syllable awareness were unique predictors of literacy skills in the control group only. Conclusions: Both early language delay and familial risk strongly overlap with subsequent dyslexia in Chinese children. Overall, rapid automatized naming and morphological awareness are relatively strong correlates of developmental dyslexia in Chinese; visual skill and phonological awareness may also be uniquely associated with subsequent literacy development in at-risk and typically developing children, respectively. Keywords: Language impairment, genetic risk, morpho- logical awareness, rapid automatized naming.
Although several large-scale research studies have now established clear cognitive correlates of dyslexia in Chinese children from mid-primary school on (e.g., Chan, Ho, Tsang, Lee, & Chung, 2006; Chung et al., 2008; Ho, Chan, Lee, Tsang, & Luan, 2004; Ho, Chan, Chung, Lee, & Tsang, 2007; Shu, McBride- Chang, Wu, & Liu, 2006), one of the most important questions for clinicians is how to identify children at-risk for reading difficulties as early as possible. The earlier such children are diagnosed, the more time there is for systematic and effective interven- tion (Hulme & Snowling, 2009; Shaywitz, 2003). In Hong Kong, where children generally begin formal instruction in literacy by about the age of 3.5 years, early predictors of dyslexia are particularly impor- tant. Wong, Kidd, Ho, and Au (2010) demonstrated that approximately 43% of Hong Kong Chinese children ages 6 to 11 years old with a prior history of early language delays manifested dyslexia, a similar pattern to that found in alphabetic languages. How- ever, that study was retrospective in nature, involv-
Conflict of interest statement: No conflicts declared.
ing children from a wide-ranging age group. A group with a narrower age range would add to these findings.
McBride-Chang et al. (2008) examined early cog- nitive correlates of Hong Kong Chinese children who were at risk for reading difficulties either because of early language delays as diagnosed by a clinician or physician or because of a genetic risk from an older sibling having already manifested dyslexia. Com- pared to a control group without any prior risks for dyslexia, children with a genetic risk for dyslexia showed particular difficulties in lexical tone detec- tion, morphological awareness, and Chinese word reading, whereas the language delayed group per- formed more poorly across all tasks administered.
The present study was a follow-up of the at-risk children tested by McBride-Chang et al. (2008). We had three questions related to these groups. First, across the two at-risk groups, what percentage of each manifested an official diagnosis of dyslexia two years later? In addition, we looked at the extent to which the children with dyslexia from the at-risk sibling group as compared to the language delayed
Ó 2010 The Authors. Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA 02148, USA
group had initially differed across cognitive and reading tasks. Second, how did those with the diag- nosis of dyslexia differ from those at-risk children who did not have dyslexia across the cognitive and reading skills administered at time 1? Third, which of these cognitive skills were most predictive of the three literacy skills tested to make up the dyslexia diagnosis at time 2, i.e., untimed word reading, one- minute word reading, and dictation? As demon- strated previously for English-speaking children (e.g., Snowling, Gallagher, & Frith, 2003), all of these questions are important for understanding how best to identify and perhaps even how to approach the issue of remediation of those Chinese children at risk for dyslexia as early as possible.
Method
Participants
In the first phase of our cohort study, 47 children at risk for dyslexia (29 boys and 18 girls; aged 5.06 years) were referred by pediatricians from the Child Assess- ment Centre (a sector under the Department of Health in Hong Kong) when they were in kindergarten. The at-risk children came from one of two groups: (1) the familial risk group, which included 26 children who had at least one elder brother or sister diagnosed with dyslexia; (2) the language delayed group, which con- tained 21 children who were first identified by parents or teachers as having language development difficulties and were thereafter tested on language/intellectual scales, with the resulting diagnosis of a language problem in one or more areas of language delay (see McBride-Chang et al., 2008 for details). Children with profound behavioural problems, such as autism, hyperactivity, or significant developmental delay, were excluded.
In the second phase of testing, these at-risk children were 7.3 years of age on average and had finished one year of primary school. They were then tested across the three literacy sub-tests of the Hong Kong Test of Spe- cific Learning Difficulties (HKT-SpLD; Ho, Chan, Tsang, & Lee, 2000). Raw scores were first converted to scaled scores base on the local norms, and the scaled scores ranged from 1 to 19, with 10 being the mean. Here, we refer to those 26 children who attained a literacy com- posite scaled score of 7 or below as the dyslexic group, while the other 21 at-risk children who scored at 8 or above on this battery were considered the non-dyslexic group (e.g., Ho et al., 2004).
Our control group consisted of 47 normally achieving children who were 5.08 years of age at time 1, matched to the at-risk group by age, gender, mother’s education level, and nonverbal IQ (assessed by Raven’s Standard Progressive Matrices; Raven, 1956). They were selected from a sample of 163 children who were taking part in an ongoing longitudinal study of language development coordinated by the first author. In the follow-up testing, all children in the control group scored higher than 7 on a composite of the HKT-SpLD literacy subtests. The average mean scaled scores on each of the three literacy subtests for the three groups of children are shown in Table 1.
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Early predictors of dyslexia in Chinese children 205
Table 1 Means and standard deviations of all tasks at times 1 and 2 controlling for age of three groups of participants, and F-value for univariate test of group differences Mean (SD)
At-risk group Non-dyslexic (N = 21)
Measures (Maximum Score)
Age (Time 2) Time 1 variables
Nonverbal intelligence (24) Visual-spatial relationships (15) Rapid number naming Morphological awareness (20) Syllable deletion (15)
Tone detection (48)
Chinese character recognition (211) Time 2 variables (Scaled scores)
Chinese word reading+ One-minute reading+ Chinese word dictation+
Dyslexic (N = 26) Mean (SD) Range 88.78 (4.80) 80–98
11.62 (2.24) 7–16 8.62 (4.45) 1–15
27.86 (7.66) 17.06–52.89 6.24 (3.63) 0–14 8.88 (4.14) 0–15
33.86 (5.25) 21–47 23.62 (14.05) 0–49
5.23 (2.10) 1–8 6.08 (1.55) 4–10 4.00 (2.02) 1–8
Mean (SD) Range 85.58 (3.08) 82–92
11.19 (2.34) 7–16 9.81 (4.63) 0–15
26.08 (15.63) 15.90–88.84 7.30 (3.66) 3–17 9.33 (4.80) 0–15
32.50 (5.20) 25–48 33.45 (14.66) 8–74
11.43 (2.34) 8–16 10.00 (2.07) 7–16 7.95 (1.88) 4–11
Control group (N = 47) Mean (SD)
90.85 (7.41)
11.13 (1.92) 10.19 (4.57) 20.05 (6.35)
8.98 (3.96) 11.09 (3.96) 34.87 (5.24) 51.85 (27.94)
12.40 (3.16) 12.53 (3.18) 10.32 (3.48)
Range 79–102
8–15
0–15 8.70–43.94 2–17 0–15
23–43 18–144
F(2, 91) 5.65**
.84 1.04
5.52** 3.73* 1.78 1.06
12.73***
Pairwise comparisons by LSD RD = RN, RD = CG, RN < CG
RD = RN, RD = CG, RN = CG^ RD=RN,RD=CG,RN=CG RD=RN,RD>CG,RN>CG RD=RN,RD<CG,RN=CG RD=RN,RD=CG,RN=CG RD=RN,RD=CG,RN=CG RD=RN,RD<CG,RN<CG
RD<RN,RD<CG,RN=CG RD<RN,RD<CG,RN<CG RD<RN,RD<CG,RN<CG
5–19 60.94*** 6–19 52.46*** 3–16 40.39***
^RD = at-risk and dyslexic; RN = at-risk and non-dyslexic; CG = control group; *p < .05., **p < .01 ***p < .001. + The means of the three SpLD literacy subtests at T2 were expressed in age-graded scaled scores (HKT-SpLD manual, 2000).
206 Catherine McBride-Chang et al. Procedure
The first testing phase (June to July 2006) was carried out when children were in kindergarten. All children were assessed on six literacy/cognitive tasks at this time. About two years later (July to October 2008), they were contacted again to be assessed for dyslexia. For the dyslexic-risk group, the testing at both times was done individually with children at a Hong Kong Child Assessment Centre by pediatricians, clinical psycholo- gists, and undergraduate psychology majors with prior training. The control group was tested as part of an ongoing longitudinal study by trained undergraduate psychology majors at the children’s homes. Informed consent was obtained from the parents before each testing phase.
Measures
The following six cognitive/literacy tasks were admin- istered to all participants at age 5.
Chinese character recognition. This task required children to read aloud 27 one-character and 34 two- character words. Testing terminated when the children failed to correctly read 10 consecutive items. For those children who were able to move on to the last item without reaching ceiling, the Chinese word reading subtest of the Hong Kong Test for Specific Learning Disabilities (HKT-SpLD; Ho et al., 2000) was then administered to further assess children’s reading abil- ities on more difficult items. The maximum score one could get across the entire assessment was 211 (61 for the first task and 150 for the second task). The internal consistency reliability of this whole word reading task was .96.
Visual-spatial skill. On each of the 16 items from the Visual-Spatial Relationships subtest of Gardner’s (1996) Test of Visual-Perceptual Skills (non-motor) Revised, children were required to select one figure from among a set of five figures with a different directionality from the others. The highest total score possible for this task was 15 (the first item was a test trial and was not counted). When the children gave four incorrect answers out of five consecutive questions, testing ceased. This task yielded an internal consistency reliability of .89.
Rapid number naming. Children were presented with a sheet of paper containing five rows of arabic numerals; each row comprised the same five digits (5, 4, 3, 1, 8) arranged in different orders. Before the testing started, the experimenters first ensured that the chil- dren could name the 5 numbers accurately. After that, the children were asked to identify aloud the 25 num- bers on the sheet in a specified order as quickly as possible. They performed the task twice, and both trials were timed using a stop-watch in order to obtain an average time. The test–retest reliability was .92 for this measure.
Morphological awareness. On this task, used in past studies (e.g., McBride-Chang et al., 2005, 2008)
to tap children’s lexical compounding abilities, chil- dren were asked to combine familiar morphemes to produce linguistically sensible compound nonwords for describing novel objects. For example, one item was: ‘When oil is made of peanuts, we call it peanut oil. What should we call it if the oil is made of mushrooms?’ (The answer should be ‘mushroom oil’.) This task taps structural knowledge within one’s lan- guage. The individual morphemes (e.g., mushroom; oil) are well known by these children. However, their primary task is to combine them in new ways for an efficient and sensible response. For example, ‘oil mushroom’ would be incorrect because its meaning is completely different from ‘mushroom oil’. All 20 items (together with the two example items at the beginning) were orally administered using a two-sentence sce- nario as in the above example. In the trial scenarios (but not the testing scenarios), pictures were pre- sented for assisting children in understanding the task requirements. Experimenters were instructed to go through all 20 items regardless of children’s per- formance. The task’s internal consistency reliability was .72.
Syllable deletion. For this task, across each of the 15 items (including two practice trials), experimenters said a three-syllable word in Cantonese, and the chil- dren were instructed to repeat the word orally with one of the three syllables omitted (either the first, middle, or last). One sample item is the following: ‘Please say/ fu6 can1 zit3/(meaning Father’s Day) without/zit3/.’ The correct answer is ‘/fu6 can1/’ (meaning father). The ceiling level was reached when incorrect answers were given for five consecutive items. An internal consistency reliability of .90 was obtained for this scale.
Tone detection.
By referring to Ciocca and Lui’s tone awareness task (2003), we selected two Cantonese monosyllables ‘/ji/and/fu/’ and made use of their six respective tones as stimuli. A total of twelve target syllables, including/ji1/ (clothing),/ji2/ (chair), /ji3/ (first character of spaghetti),/ji4/ (son),/ji5/ (ear),/ji6/ (two),/fu1/ (skin),/fu2/ (tiger),/ fu3/ (trousers),/fu4/ (symbol),/fu5/ (woman), and/fu6/ (father), were each pictorially represented by a concrete object as indicated above. In the training session, children learned to identify the syllable cor- responding to each of the twelve pictures. A booklet containing three trials and 48 test items was used in the testing phase, and all the verbal stimuli were played using a Sony mini-disc player. For each item, children were presented with two pictures with each of them derived from the same monosyllable (either/ji/ or/fu/), but having different tones (e.g./ji2/ and/ ji5/ ). After listening to the target syllable through the mini-disc player, children were asked to choose from two pictures the one that corresponded to the target syllable. The 48 testing items were designed such that each of the 12 tones was the target word for 4 items, across which they were paired with four other different tones (12 · 4 = 48). The internal consistency
reliability of this task was .66. Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Ó 2010 The Authors
Hong Kong Test of Specific Learning Difficulties in Reading and Writing
The three literacy subtests of the HKT-SpLD (Ho et al., 2000) were administered to all participants at age 7 only. The first subtest was Chinese word reading con- taining 150 two-character words arranged in order of increasing difficulty. Children were instructed to read the words aloud one by one until they failed to read 15 consecutive words. The One-minute reading subtest tests children’s reading fluency; here, children are given one minute to read 90 two-character Chinese words as quickly and as accurately as possible. Their total score is the number of words correctly read within the 60-second time period. The third subtest, Chinese word dictation, requires children to write 48 two-character words. One correctly written character yielded one mark, and the testing ended when the child scored a zero for both characters across 8 consecutive words. The raw scores attained in the three subtests were converted to age-graded scaled scores (HKT-SpLD manual; Ho et al., 2000). Following the manual, those who scored a 7 or below averaged across the three subtests were identified as dyslexic.
Results
Among the 47 preschool children considered at risk for reading problems at time 1, 26 of them were confirmed to be dyslexic in the follow-up testing while the other 21 were identified as non-dyslexic. Among those diagnosed as dyslexic, 13 were from the genetically at-risk group, whereas the other 13 were from the language delayed group. Thus, there was a 50% rate of dyslexia among those children who had one older brother or sister with dyslexia (genetic risk) and approximately a 62% rate of dyslexia among those who had manifested an early language delay. These rates are in line with those from other previous studies of both genetic (Gilger, Pennington, & DeFries, 1991; Pennington, 1991; Snowling et al., 2003) and language delay (Aram & Hall, 1989; Bashir & Scavuzzo, 1992) risk. When we compared those 13 children each in the genetic and language delayed groups, respectively, who had dyslexia on all tasks administered at times 1 and 2, we found no differences in performances on any cognitive or lit- eracy tasks across time between them. Thus, for all subsequent analyses, these two groups were com- bined to form a larger group with dyslexia. All chil- dren who had been at risk for dyslexia but did not manifest the disorder at time 2 were then labeled the non-dyslexic group.
About eight months following time 2 testing, we telephoned parents of as many at-risk children as possible to determine whether any of them were receiving any special interventions at school. Of the 21 out of 26 dyslexic children of families who could be contacted (4 families were unreachable after repeated calls), 10 were currently receiving some therapy at school, and 1 additional child had had
some therapy previously but was no longer receiving it. Of the 19 non-dyslexic children whose families were reachable by phone, 7 of them had previously experienced some type of therapy at school; of these 7, 4 were currently still receiving it. Most of the intervention therapies were focused on articulation and speech organization; a minority incorporated orthographic knowledge training as well. Beyond these language-related therapies, some of the chil- dren additionally received occupational therapy, focused more specifically on fine motor skills and attention. Each therapy session at school lasted from 15 minutes to 1.5 hours; some occurred monthly, and others occurred weekly.
Our focus shifted next to a comparison of the performances among the three groups (i.e., dyslexic, non-dyslexic, and controls) across the six metalin- guistic tasks administered two years previously. The three groups differed in terms of age at time 2 (with the mean ages for at-risk dyslexic, at-risk non-dys- lexic, and control children being 88.8, 85.6 and 90.9 months, respectively, F(2,91) = 5.65, p < .05). The children who were at risk for dyslexia but non-dys- lexic differed in age from the other two groups, which did not differ from one another. However, the groups differed neither in mothers’ education level (F(2,91) = .28, n.s.) nor in nonverbal intelligence (F(2,91) = .47, n.s.) at time 1. Therefore, the task performances were compared using ANCOVA, with age at time 2 as the only covariate. Results of ANCOVAs (Table 1) showed that there was a significant overall group difference across all experimental tasks except for Visual-Spatial Relationships, syllable deletion, and tone detection.
The performances of the control children were significantly better than those in both at-risk groups (dyslexic and not) in rapid number naming, F(2, 90) = 5.52, p < .01, partial g2 = .11. In contrast, for the morphological awareness tasks, the control group scored significantly higher than those 26 children with dyslexia only, F(2, 91) = 3.73, p < .05, partial g2 = .08. For Chinese character recognition at time 1, the at-risk non-dyslexic performed as poorly as the at-risk dyslexic group and worse than the control group (F(2, 91) = 12.73, p < .001, partial g2 = .22).
Apart from looking at group differences, we per- formed correlational and regression analyses to explore the role of the metalinguistic measures as early predictors of children’s subsequent reading and writing outcomes. Because there was some evidence that the at-risk non-dyslexic group scored between the at-risk dyslexic and control groups (i.e., significantly higher than the dyslexics but lower than the controls) and also for simplicity, correlations are shown in Table 2 separately only for the at-risk and control groups, though the patterns of correlations for the dyslexic and non-dyslexic at-risk groups were similar. Because mothers’ education levels were not significantly associated with any of the perfor- mance measures, this variable was not included in
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Early predictors of dyslexia in Chinese children 207
208
Catherine McBride-Chang et al.
correlational or any subsequent analyses. However, because age was significantly associated with some variables, partial correlations among all variables are shown in Table 2 with age statistically con- trolled.
We then examined the extent to which we could predict individual literacy skills two years later, based on time 1 cognitive tasks, in the entire group com- bined in order to maximize statistical power, given the relatively large number of variables included. We included sex in these analyses because for dictation, but not for the other variables, girls significantly outperformed boys. Given inter-correlations among predictors, we report sr2 (in addition to b) in Table 3 to reflect the unique variance explained by each pre- dictor after accounting for the variances contributed by the other predictors combined.
Without controlling for Chinese word reading at time 1, between 26% and 31% of the variance could be explained by these variables across groups at time 2. The strongest correlate of word reading was mor- phological awareness; morphological awareness was also marginally (p < .07) significantly associated with one-minute reading and dictation. Speeded naming was a unique predictor of one-minute reading and marginally associated with dictation. Age tended to be significantly negatively associated with these literacy skills because we used scaled, rather than raw, scores to predict these variables. (Because some of these assessments were made by busy clinicians, raw scores were not always available for these analyses.) Age was already taken into account in the scaled scores, but it was included here because of its positive associations with some of the cognitive variables (i.e., morphological awareness, syllable deletion, and Visual-Spatial Relationships) at time 1. Overall, these analyses indicate that Hong Kong Chinese children’s skills in speeded naming and morphological aware- ness may be among the most sensitive predictors of subsequent performance on literacy skills. Syllable deletion, tone detection, and visual skills, though correlated with subsequent literacy skills, were not unique predictors of subsequent Chinese word reading in this combined group.
We also carried out these analyses for the three literacy skills with the autoregressive effects of Chi- nese word reading at time 1 included. In these anal- yses, morphological awareness remained a unique predictor of subsequent word reading only. In con- trast, only speeded naming was a unique predictor of subsequent speeded reading. None of the cognitive tasks were unique predictors of dictation skill apart from early word reading itself. These analyses indi- cate that even for very young Hong Kong Chinese children, Chinese word reading itself, if available, is the best predictor of subsequent literacy skills.
Another way to analyze these data is to determine which variables uniquely explain variance in the lit- eracy tasks using stepwise (forward) regression analyses. We did this separately for the control and
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Table 2 Partial correlations among time 1 metalinguistic measures and time 2 literacy outcomes (word reading, one-minute reading and word dictation) with time 1 age controlled Control (N = 47)
At-risk (N = 47) Nonverbal intelligence
Visual-spatial relationships Rapid number naming Morphological awareness Syllable deletion
Tone detection Chinese character recognition Chinese word reading (T2) One-minute reading (T2) Chinese word dictation (T2)
Rapid Nonverbal Visual-spatial number Morphological
intelligence relationships naming awareness .22 .12 .07
.61 ).01 .11 ).08 ).46 ).02
.39 .31 ).34 .13 .35 ).48 .28 .36 .30 ).16 .33 .16 .16 ).21 .55 .16 .33 ).14 .33 .13 .33 ).39 .34 .16 .34 ).18 .22
Chinese Chinese Syllable Tone character word
deletion detection recognition reading (T2)
.13 .18 .11 .12
.31 .38 .21 .05 ).27 ).27 ).22 ).18 .28 .57 .25 .34 .40 .26 .16 .28 .56 .42 .37 .20 .55
.19 .08 .47 .29 .04 .32 .67 .08 ).01 .22 .72
One-minute reading (T2)
.18 ).02 ).49 .12 .09 .37 .45 .57
.56
Chinese word dictation (T2)
.15
.05 ).32 .19 .43 .28 .40 .57 .50
Note: Correlations of magnitude.31 or higher were significant at p < .05.
Early predictors of dyslexia in Chinese children 209 Table 3 Final standardized beta weights explaining three literacy outcomes at time 2 from metalinguistic measures at time 1
Age ).38 Sex .05
Nonverbal intelligence Visual-spatial relationships Rapid number naming Morphological awareness Syllable deletion
Tone detection Chinese word reading R2 (df)
).23 .03 .06 ).06 ).39 .07 ).06 ).03 .48 .46
).20* ).22 .03 .06 .05 .03
).05 ).04 ).33*** ).43 .06 .21 ).05 ).00 ).03 .09
.40*** (9, 82) .31
).19* .06 .16 .03 .04
).26** ).29 .15+ .18
).25** .17+
With T1 read- ing
b sr2
Word reading
One-minute reading
Dictation
Without T1 reading
b sr2
).32** .07
).05 .07 ).11
.30** .02 .05
.55 .45*** .48 (9, 82) .28 (8, 84)
With T1 read- ing
b sr2
Without T1 reading
b sr2
With T1 read- ing
b sr2
Without T1 reading
b sr2
).33*** ).37 .04 .07 ).02 ).02 ).06 .08 .06 .09 ).07 ).06 ).13 .21 .17* .36 ).04 ).04 .02 ).07 ).06 .07
).30 ).03 .01
+p < .10, *p < .05, **p < .01, ***p < .001. sr2 represents the part correlation, i.e., the correlation between the predictor and the criterion variable with other predictors statistically controlled.
at-risk groups. (Because the results for the whole combined sample mirrored those shown in Table 3, they are not reviewed here.) For the control group, with word reading from time 1 included, only time 1 word reading uniquely emerged as a correlate of word reading at time 2 (total R-squared = .26), only speeded naming and time 1 word reading emerged as predictors of One-Minute Reading (total R-squared = .33), and only syllable deletion emerged as a unique correlate of dictation (total R-squared = .16). In contrast, the unique predictors for the at-risk chil- dren were age, word reading at time 1, and Visual- Spatial Relationships for word reading at time 2 (total R-squared = .45), age and speeded naming for One-Minute Reading (total R-squared = .20), and age and Visual-Spatial Relationships for dictation (total R-squared = .24). Patterns remained the same when word reading at time 1 was excluded in the analyses for the control and at-risk groups separately, with one exception: only tone detection was a unique predictor of word reading in the control group (total R-squared = .16).
Discussion
The present study was a follow-up on previous work (McBride-Chang et al., 2008) seeking to identify those Hong Kong Chinese children at greatest risk for dyslexia at a very young age based on a battery of cognitive tasks. We demonstrated that 50% of those at genetic risk for reading difficulties and 62% of those with a diagnosed language delay at age 5 subsequently manifested dyslexia in this sample. The genetic and language delayed groups did not differ from one another in severity of dyslexia. When groups of children who had and had not subse- quently manifested dyslexia at age 7 were compared to one another and to a control group, neither the visual spatial skill nor the lexical tone or syllable
deletion tasks were clear unique predictors of sub- sequent reading difficulties. However, measures of speeded naming and morphological awareness did distinguish the children who were ultimately diag- nosed as dyslexic from a control group. Moreover, only these two measures emerged as unique pre- dictors in the entire group of one or more subsequent literacy skills. In addition, very early word reading was the best measure for distinguishing among children who subsequently manifest the diagnosis of dyslexia.
These results are practically important. They underscore the overlap between language delay and dyslexia found in previous work in Hong Kong (Wong et al., 2010). In addition, they highlight the impor- tance of going beyond phonological sensitivity mea- sures only to diagnose subsequent reading difficulties in Chinese children. As demonstrated in correlational studies of older Chinese children with dyslexia, both speeded naming (Ho & Lai, 1999) and morphological awareness (Chung et al., 2008; Shu et al., 2006; Wong et al., 2010) are key predictors of reading difficulties in Chinese children. Moreover, given the unique educational structure of Hong Kong, a city in which formal literacy instruction begins very early (e.g., Cheung & Ng, 2003), it is equally vital to note that Chinese word reading itself at age 5 tends to be the best predictor of subsequent reading difficulties in older children. At the same time, however, reading itself is strongly influenced by explicit teaching. Therefore, clinicians should use their own judgment in assessing the importance or weight assigned to early word reading in predicting subsequent reading difficulties in at-risk groups.
Theoretically, these results also demonstrate the broad array of metalinguistic and cognitive skills that are potentially important for learning to read Chinese. Although phonological sensitivity is clearly developmentally vital for learning to read Chinese, the unreliability of phonological cues to Chinese
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
).37*** .17+
).00 .07
(8, 84)
.03
.01 ).17 ).14 .11 .08 .08 .07 ).09 ).07
.44 .37*** .39 (9, 82)
.01 .01
.02 .02 ).21 ).18+ .23 .19+
.14 .11 .02 .02
.26 (8, 84)
210 Catherine McBride-Chang et al.
words, in addition to the relatively easy phonological structure of the Chinese language, makes it essential to look at other indicators of reading difficulties. Morphological awareness may be particularly helpful in early acquisition of Chinese literacy skill. In addition, fluency in identifying printed symbols, in this case in the form of simple numbers, is a strong correlate of subsequent reading, particularly spee- ded reading.
At the same time, forward stepwise regression analyses by group also indicated that all of the measures included at time 1 were uniquely asso- ciated with at least one time 2 literacy skill in either the control or at-risk group. For example, both the syllable awareness and tone detection measures explained unique variance in subsequent literacy skill in the control group. Thus, within this group, with, almost by definition, greater variability across all literacy tasks at time 2 as compared to those who were initially at risk for reading difficulties, phonological sensitivity measures were significant predictors of subsequent reading and writing. In contrast, in the at-risk group, early visual, but not phonological sensitivity, skills were uniquely asso- ciated with subsequent dictation and untimed word reading, suggesting that visual skills may be useful, perhaps somewhat independently of (phonological or morphological) metalinguistic skills, for Chinese literacy acquisition among at-risk readers. Future work should consider the possibility that the route to Chinese literacy acquisition may include multiple dimensions which may differ somewhat across those with and without risks for reading difficulties.
Importantly, those at-risk children who ultimately did not manifest dyslexia two years later neverthe- less performed significantly more poorly than the control group on two of the three literacy skills tested at time 2, though they had differed from the control group at time 1 only on speeded naming and word reading. This pattern is similar to one demonstrated in a longitudinal study of English-speaking children at familial risk for dyslexia and adds weight to the argument (Snowling et al., 2003) that solid language skills may help children at risk for dyslexia to com- pensate for some reading difficulties. At the same
time, the fact that this at-risk group was initially slower on the rapid naming task also suggests that automatization of literacy skills may be a continuing challenge for them.
The present study was limited in the measures it included. As noted previously (McBride-Chang et al., 2008), ours was not an exhaustive group of cognitive tasks. For example, paired associate learning might have been a good measure of early learning to include (Li, Shu, McBride-Chang, Liu, & Xue, 2009). Future studies might increase the number of mea- sures tested to predict subsequent reading difficul- ties. In addition, there were some age-diagnosis interactions. Indeed, the children who had been initially at risk for dyslexia but did not manifest the disorder were significantly younger than were those of the same grade level who were diagnosed with dyslexia. Criteria for the diagnosis of dyslexia are stricter for older children.
Despite these limitations, however, this study has been important in at least two ways. First, we have examined longitudinally in a prospective study the overlap between genetic risk, as well as early lan- guage delay, and subsequent dyslexia in Chinese. This overlap is considerable and adds to the growing understanding of how important early screening is for later reading problems. Second, we have isolated speeded naming and morphological awareness, apart from reading itself, as two cognitive tasks that appear to be potentially clinically useful in identify- ing very young Chinese children at risk for sub- sequent reading difficulties.
Acknowledgements
We are grateful to the Research Grants Council of the Hong Kong Special Administrative Region (project reference #448907) for support of this research.
Correspondence to
Catherine McBride-Chang, Psychology Department, The Chinese University of Hong Kong, Shatin, Hong Kong, China; Tel: 852-26096576; Email: cmcbride@ psy.cuhk.edu.hk
Key points
• The rate of dyslexia at age 7 in Chinese children with a language delay at age 5 was 62% and with a genetic risk for dyslexia at age 5 was 50%.
• The cognitive tasks at age 5 that best distinguished Chinese children with and without dyslexia at age 7 and that predicted subsequent literacy skills across all children were rapid automatized naming and morphological awareness. At the same time, word reading at time 1 itself was the best predictor of sub- sequent reading difficulties in this sample.
• Within the control group only, syllable deletion and tone sensitivity were significantly associated with subsequent literacy skills; among the at-risk group, visual skills were significantly associated with sub- sequent literacy skills.
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
References
Aram, D.M., & Hall, N.C. (1989). Longitudinal follow-up of children with preschool communication disorders: Treatment implications. School Psychology Review, 18, 487–501.
Bashir, A.S., & Scavuzzo, A. (1992). Children with lan- guage disorders: Natural history and academic success. Journal of Learning Disabilities, 25, 53–65.
Chan, D.W., Ho, C.S.-.H, Tsang, S.-M., Lee, S.-H., & Chung, K.K.H. (2006). Exploring the reading-writing connection in Chinese children with dyslexia in Hong Kong. Reading and Writing: An Interdisciplinary Journal, 19, 543–561.
Cheung, H., & Ng, L. (2003). Chinese reading development in some major Chinese societies: An introduction. In C. McBride-Chang, & H.-C. Chen (Eds.), Reading develop- ment in Chinese children (pp. 3–17). Westport, CT: Praeger Press.
Chung, K.K.H., McBride-Chang, C., Wong, S.W.L., Che- ung, H., Penney, T.B., & Ho, C.S.-H. (2008). The role of visual and auditory temporal processing for Chinese children with developmental dyslexia. Annals of Dys- lexia, 58, 15–35.
Ciocca, V., & Lui, J.Y.K. (2003). The development of the perception of Cantonese lexical tones. Journal of Multi- lingual Communication Disorders, 1, 141–147.
Gardner, M.F. (1996). Test of visual-perceptual skills (non- motor). Burlingame, CA: Psychological and Educational Publications.
Gilger, J.W., Pennington, B.F., & DeFries, J.C. (1991). Risk for reading disability as a function of parental history in three family studies. Reading and Writing: An Interdisci- plinary Journal, 3, 299–313.
Ho, C.S.-H., Chan, D.W., Chung, K.K.H., Lee, S.-H., & Tsang, S.-M. (2007). In search of subtypes of Chinese developmental dyslexia. Journal of Experimental Child Psychology, 97, 61–83.
Ho, C.S.-H., Chan, D.W.-O., Lee, S.-H., Tsang, S.-M., & Luan, H.V. (2004). Cognitive profiling and preliminary subtyping in Chinese developmental dyslexia. Cognition, 91, 43–75.
Ho, C.S.-H., Chan, D.W.-O., Tsang, S.-M., & Lee, S.-H. (2000). The Hong Kong Test of Specific Learning Difficul- ties in Reading and Writing. Hong Kong: Hong Kong Specific Learning Difficulties Research Team.
Ho, C.S.-H., & Lai, D.N.-C. (1999). Naming-speed deficits and phonological memory deficits in Chinese develop- mental dyslexia. Learning and Individual Differences, 11, 173–186.
Hulme, C., & Snowling, M.J. (2009). Developmental disor- ders of language learning and cognition. Chichester: Wiley-Blackwell.
Li, H., Shu, H., McBride-Chang, C., Liu, H.-Y., & Xue, J. (2009). Paired associate learning in Chinese children with dyslexia. Journal of Experimental Child Psychology, 103, 135–151.
McBride-Chang, C., Cho, J.-R., Liu, H., Wagner, R.K., Shu, H., Zhou, A., Cheuk, C.S.-M., & Muse, A. (2005). Changing models across cultures: Associations of pho- nological and morphological awareness to reading in Beijing, Hong Kong, Korea, and America. Journal of Experimental Child Psychology, 92, 140–160.
McBride-Chang, C., Lam, F., Lam, C., Doo, S., Wong, S.W.L., & Chow, Y.Y.Y. (2008). Word recognition and cognitive profiles of Chinese pre-school children at-risk for dyslexia through language delay or familial history of dyslexia. Journal of Child Psychology and Psychiatry, 49, 211–218.
Pennington, B.F. (1991). Diagnosing learning disorders: A neuropsychological framework. New York: Guilford Press.
Raven, J.C. (1956). Coloured progressive matrices. Oxford: Psychologists Press.
Shaywitz, S.E. (2003). Overcoming dyslexia. New York: Alfred A. Knopf.
Shu, H., McBride-Chang, C., Wu, H., & Liu, H. (2006). Understanding Chinese developmental dyslexia: Mor- phological awareness as a core cognitive construct. Journal of Educational Psychology, 98, 122–133.
Snowling, M.J., Gallagher, A., & Frith, U. (2003). Family risk of dyslexia is continuous: Individual differences in the precursors of reading skill. Child Development, 74, 358–373.
Wong, A.M.Y., Kidd, J., Ho, C.S.-H., & Au, T.K.-F. (2010). Characterizing the overlap between SLI and dyslexia in Chinese: The role of phonology and beyond. Scientific Studies of Reading, 14, 30–57.
Manuscript accepted 24 May 2010
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Early predictors of dyslexia in Chinese children 211
It is a great concern that young children in China will be Dysexia
I believe a young child's development should be assessed. The reason being, the teacher has to have a starting point. This will help the teacher understand were the child is in terms of academics. Teachers refer to this as a base-line.
The additional ideas and suggestions I have to help children with problems be dysexia, try having the child tested earlier than their 5 birthday.
Journal of Child Psychology and Psychiatry 52:2 (2011), pp 204–211 doi:10.1111/j.1469-7610.2010.02299.x
Early predictors of dyslexia in Chinese children: familial history of dyslexia, language delay, and cognitive profiles
Catherine McBride-Chang,1 Fanny Lam,2 Catherine Lam,2 Becky Chan,2 Cathy Y.-C. Fong,3 Terry T.-Y. Wong,1 and Simpson W.-L. Wong4
1The Chinese University of Hong Kong, China; 2Child Assessment Service, Department of Health, Hong Kong, China; 3The University of Hong Kong, China; 4University of Oxford, UK
Background: This work tested the rates at which Chinese children with either language delay or familial history of dyslexia at age 5 manifested dyslexia at age 7, identified which cognitive skills at age 5 best distinguished children with and without dyslexia at age 7, and examined how these early abilities predicted subsequent literacy skills. Method: Forty-seven at-risk children (21 who were ini- tially language delayed and 26 with familial risk) and 47 control children matched on age, IQ, and mothers’ education were tested on syllable awareness, tone detection, rapid automatized naming, visual skill, morphological awareness, and word reading at age 5 and subsequently tested for dyslexia on a standard Hong Kong measure at age 7. Results: Of those with an early language delay, 62% subsequently manifested dyslexia; for those with familial risk, the rate of dyslexia was 50%. Those with dyslexia were best distinguished from those without dyslexia by the age-5 measures of morphological awareness, rapid automatized naming, and word reading itself; other measures did not distinguish the groups. In a combined regression analysis across all participants, morphological awareness uniquely explained word reading accuracy and rapid automatized naming uniquely explained timed word reading at age 7, with all other measures statistically controlled. Separate stepwise regression analyses by group indicated that visual skill uniquely explained subsequent literacy skills in the at-risk group only, whereas tone and syllable awareness were unique predictors of literacy skills in the control group only. Conclusions: Both early language delay and familial risk strongly overlap with subsequent dyslexia in Chinese children. Overall, rapid automatized naming and morphological awareness are relatively strong correlates of developmental dyslexia in Chinese; visual skill and phonological awareness may also be uniquely associated with subsequent literacy development in at-risk and typically developing children, respectively. Keywords: Language impairment, genetic risk, morpho- logical awareness, rapid automatized naming.
Although several large-scale research studies have now established clear cognitive correlates of dyslexia in Chinese children from mid-primary school on (e.g., Chan, Ho, Tsang, Lee, & Chung, 2006; Chung et al., 2008; Ho, Chan, Lee, Tsang, & Luan, 2004; Ho, Chan, Chung, Lee, & Tsang, 2007; Shu, McBride- Chang, Wu, & Liu, 2006), one of the most important questions for clinicians is how to identify children at-risk for reading difficulties as early as possible. The earlier such children are diagnosed, the more time there is for systematic and effective interven- tion (Hulme & Snowling, 2009; Shaywitz, 2003). In Hong Kong, where children generally begin formal instruction in literacy by about the age of 3.5 years, early predictors of dyslexia are particularly impor- tant. Wong, Kidd, Ho, and Au (2010) demonstrated that approximately 43% of Hong Kong Chinese children ages 6 to 11 years old with a prior history of early language delays manifested dyslexia, a similar pattern to that found in alphabetic languages. How- ever, that study was retrospective in nature, involv-
Conflict of interest statement: No conflicts declared.
ing children from a wide-ranging age group. A group with a narrower age range would add to these findings.
McBride-Chang et al. (2008) examined early cog- nitive correlates of Hong Kong Chinese children who were at risk for reading difficulties either because of early language delays as diagnosed by a clinician or physician or because of a genetic risk from an older sibling having already manifested dyslexia. Com- pared to a control group without any prior risks for dyslexia, children with a genetic risk for dyslexia showed particular difficulties in lexical tone detec- tion, morphological awareness, and Chinese word reading, whereas the language delayed group per- formed more poorly across all tasks administered.
The present study was a follow-up of the at-risk children tested by McBride-Chang et al. (2008). We had three questions related to these groups. First, across the two at-risk groups, what percentage of each manifested an official diagnosis of dyslexia two years later? In addition, we looked at the extent to which the children with dyslexia from the at-risk sibling group as compared to the language delayed
Ó 2010 The Authors. Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA 02148, USA
group had initially differed across cognitive and reading tasks. Second, how did those with the diag- nosis of dyslexia differ from those at-risk children who did not have dyslexia across the cognitive and reading skills administered at time 1? Third, which of these cognitive skills were most predictive of the three literacy skills tested to make up the dyslexia diagnosis at time 2, i.e., untimed word reading, one- minute word reading, and dictation? As demon- strated previously for English-speaking children (e.g., Snowling, Gallagher, & Frith, 2003), all of these questions are important for understanding how best to identify and perhaps even how to approach the issue of remediation of those Chinese children at risk for dyslexia as early as possible.
Method
Participants
In the first phase of our cohort study, 47 children at risk for dyslexia (29 boys and 18 girls; aged 5.06 years) were referred by pediatricians from the Child Assess- ment Centre (a sector under the Department of Health in Hong Kong) when they were in kindergarten. The at-risk children came from one of two groups: (1) the familial risk group, which included 26 children who had at least one elder brother or sister diagnosed with dyslexia; (2) the language delayed group, which con- tained 21 children who were first identified by parents or teachers as having language development difficulties and were thereafter tested on language/intellectual scales, with the resulting diagnosis of a language problem in one or more areas of language delay (see McBride-Chang et al., 2008 for details). Children with profound behavioural problems, such as autism, hyperactivity, or significant developmental delay, were excluded.
In the second phase of testing, these at-risk children were 7.3 years of age on average and had finished one year of primary school. They were then tested across the three literacy sub-tests of the Hong Kong Test of Spe- cific Learning Difficulties (HKT-SpLD; Ho, Chan, Tsang, & Lee, 2000). Raw scores were first converted to scaled scores base on the local norms, and the scaled scores ranged from 1 to 19, with 10 being the mean. Here, we refer to those 26 children who attained a literacy com- posite scaled score of 7 or below as the dyslexic group, while the other 21 at-risk children who scored at 8 or above on this battery were considered the non-dyslexic group (e.g., Ho et al., 2004).
Our control group consisted of 47 normally achieving children who were 5.08 years of age at time 1, matched to the at-risk group by age, gender, mother’s education level, and nonverbal IQ (assessed by Raven’s Standard Progressive Matrices; Raven, 1956). They were selected from a sample of 163 children who were taking part in an ongoing longitudinal study of language development coordinated by the first author. In the follow-up testing, all children in the control group scored higher than 7 on a composite of the HKT-SpLD literacy subtests. The average mean scaled scores on each of the three literacy subtests for the three groups of children are shown in Table 1.
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Early predictors of dyslexia in Chinese children 205
Table 1 Means and standard deviations of all tasks at times 1 and 2 controlling for age of three groups of participants, and F-value for univariate test of group differences Mean (SD)
At-risk group Non-dyslexic (N = 21)
Measures (Maximum Score)
Age (Time 2) Time 1 variables
Nonverbal intelligence (24) Visual-spatial relationships (15) Rapid number naming Morphological awareness (20) Syllable deletion (15)
Tone detection (48)
Chinese character recognition (211) Time 2 variables (Scaled scores)
Chinese word reading+ One-minute reading+ Chinese word dictation+
Dyslexic (N = 26) Mean (SD) Range 88.78 (4.80) 80–98
11.62 (2.24) 7–16 8.62 (4.45) 1–15
27.86 (7.66) 17.06–52.89 6.24 (3.63) 0–14 8.88 (4.14) 0–15
33.86 (5.25) 21–47 23.62 (14.05) 0–49
5.23 (2.10) 1–8 6.08 (1.55) 4–10 4.00 (2.02) 1–8
Mean (SD) Range 85.58 (3.08) 82–92
11.19 (2.34) 7–16 9.81 (4.63) 0–15
26.08 (15.63) 15.90–88.84 7.30 (3.66) 3–17 9.33 (4.80) 0–15
32.50 (5.20) 25–48 33.45 (14.66) 8–74
11.43 (2.34) 8–16 10.00 (2.07) 7–16 7.95 (1.88) 4–11
Control group (N = 47) Mean (SD)
90.85 (7.41)
11.13 (1.92) 10.19 (4.57) 20.05 (6.35)
8.98 (3.96) 11.09 (3.96) 34.87 (5.24) 51.85 (27.94)
12.40 (3.16) 12.53 (3.18) 10.32 (3.48)
Range 79–102
8–15
0–15 8.70–43.94 2–17 0–15
23–43 18–144
F(2, 91) 5.65**
.84 1.04
5.52** 3.73* 1.78 1.06
12.73***
Pairwise comparisons by LSD RD = RN, RD = CG, RN < CG
RD = RN, RD = CG, RN = CG^ RD=RN,RD=CG,RN=CG RD=RN,RD>CG,RN>CG RD=RN,RD<CG,RN=CG RD=RN,RD=CG,RN=CG RD=RN,RD=CG,RN=CG RD=RN,RD<CG,RN<CG
RD<RN,RD<CG,RN=CG RD<RN,RD<CG,RN<CG RD<RN,RD<CG,RN<CG
5–19 60.94*** 6–19 52.46*** 3–16 40.39***
^RD = at-risk and dyslexic; RN = at-risk and non-dyslexic; CG = control group; *p < .05., **p < .01 ***p < .001. + The means of the three SpLD literacy subtests at T2 were expressed in age-graded scaled scores (HKT-SpLD manual, 2000).
206 Catherine McBride-Chang et al. Procedure
The first testing phase (June to July 2006) was carried out when children were in kindergarten. All children were assessed on six literacy/cognitive tasks at this time. About two years later (July to October 2008), they were contacted again to be assessed for dyslexia. For the dyslexic-risk group, the testing at both times was done individually with children at a Hong Kong Child Assessment Centre by pediatricians, clinical psycholo- gists, and undergraduate psychology majors with prior training. The control group was tested as part of an ongoing longitudinal study by trained undergraduate psychology majors at the children’s homes. Informed consent was obtained from the parents before each testing phase.
Measures
The following six cognitive/literacy tasks were admin- istered to all participants at age 5.
Chinese character recognition. This task required children to read aloud 27 one-character and 34 two- character words. Testing terminated when the children failed to correctly read 10 consecutive items. For those children who were able to move on to the last item without reaching ceiling, the Chinese word reading subtest of the Hong Kong Test for Specific Learning Disabilities (HKT-SpLD; Ho et al., 2000) was then administered to further assess children’s reading abil- ities on more difficult items. The maximum score one could get across the entire assessment was 211 (61 for the first task and 150 for the second task). The internal consistency reliability of this whole word reading task was .96.
Visual-spatial skill. On each of the 16 items from the Visual-Spatial Relationships subtest of Gardner’s (1996) Test of Visual-Perceptual Skills (non-motor) Revised, children were required to select one figure from among a set of five figures with a different directionality from the others. The highest total score possible for this task was 15 (the first item was a test trial and was not counted). When the children gave four incorrect answers out of five consecutive questions, testing ceased. This task yielded an internal consistency reliability of .89.
Rapid number naming. Children were presented with a sheet of paper containing five rows of arabic numerals; each row comprised the same five digits (5, 4, 3, 1, 8) arranged in different orders. Before the testing started, the experimenters first ensured that the chil- dren could name the 5 numbers accurately. After that, the children were asked to identify aloud the 25 num- bers on the sheet in a specified order as quickly as possible. They performed the task twice, and both trials were timed using a stop-watch in order to obtain an average time. The test–retest reliability was .92 for this measure.
Morphological awareness. On this task, used in past studies (e.g., McBride-Chang et al., 2005, 2008)
to tap children’s lexical compounding abilities, chil- dren were asked to combine familiar morphemes to produce linguistically sensible compound nonwords for describing novel objects. For example, one item was: ‘When oil is made of peanuts, we call it peanut oil. What should we call it if the oil is made of mushrooms?’ (The answer should be ‘mushroom oil’.) This task taps structural knowledge within one’s lan- guage. The individual morphemes (e.g., mushroom; oil) are well known by these children. However, their primary task is to combine them in new ways for an efficient and sensible response. For example, ‘oil mushroom’ would be incorrect because its meaning is completely different from ‘mushroom oil’. All 20 items (together with the two example items at the beginning) were orally administered using a two-sentence sce- nario as in the above example. In the trial scenarios (but not the testing scenarios), pictures were pre- sented for assisting children in understanding the task requirements. Experimenters were instructed to go through all 20 items regardless of children’s per- formance. The task’s internal consistency reliability was .72.
Syllable deletion. For this task, across each of the 15 items (including two practice trials), experimenters said a three-syllable word in Cantonese, and the chil- dren were instructed to repeat the word orally with one of the three syllables omitted (either the first, middle, or last). One sample item is the following: ‘Please say/ fu6 can1 zit3/(meaning Father’s Day) without/zit3/.’ The correct answer is ‘/fu6 can1/’ (meaning father). The ceiling level was reached when incorrect answers were given for five consecutive items. An internal consistency reliability of .90 was obtained for this scale.
Tone detection.
By referring to Ciocca and Lui’s tone awareness task (2003), we selected two Cantonese monosyllables ‘/ji/and/fu/’ and made use of their six respective tones as stimuli. A total of twelve target syllables, including/ji1/ (clothing),/ji2/ (chair), /ji3/ (first character of spaghetti),/ji4/ (son),/ji5/ (ear),/ji6/ (two),/fu1/ (skin),/fu2/ (tiger),/ fu3/ (trousers),/fu4/ (symbol),/fu5/ (woman), and/fu6/ (father), were each pictorially represented by a concrete object as indicated above. In the training session, children learned to identify the syllable cor- responding to each of the twelve pictures. A booklet containing three trials and 48 test items was used in the testing phase, and all the verbal stimuli were played using a Sony mini-disc player. For each item, children were presented with two pictures with each of them derived from the same monosyllable (either/ji/ or/fu/), but having different tones (e.g./ji2/ and/ ji5/ ). After listening to the target syllable through the mini-disc player, children were asked to choose from two pictures the one that corresponded to the target syllable. The 48 testing items were designed such that each of the 12 tones was the target word for 4 items, across which they were paired with four other different tones (12 · 4 = 48). The internal consistency
reliability of this task was .66. Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Ó 2010 The Authors
Hong Kong Test of Specific Learning Difficulties in Reading and Writing
The three literacy subtests of the HKT-SpLD (Ho et al., 2000) were administered to all participants at age 7 only. The first subtest was Chinese word reading con- taining 150 two-character words arranged in order of increasing difficulty. Children were instructed to read the words aloud one by one until they failed to read 15 consecutive words. The One-minute reading subtest tests children’s reading fluency; here, children are given one minute to read 90 two-character Chinese words as quickly and as accurately as possible. Their total score is the number of words correctly read within the 60-second time period. The third subtest, Chinese word dictation, requires children to write 48 two-character words. One correctly written character yielded one mark, and the testing ended when the child scored a zero for both characters across 8 consecutive words. The raw scores attained in the three subtests were converted to age-graded scaled scores (HKT-SpLD manual; Ho et al., 2000). Following the manual, those who scored a 7 or below averaged across the three subtests were identified as dyslexic.
Results
Among the 47 preschool children considered at risk for reading problems at time 1, 26 of them were confirmed to be dyslexic in the follow-up testing while the other 21 were identified as non-dyslexic. Among those diagnosed as dyslexic, 13 were from the genetically at-risk group, whereas the other 13 were from the language delayed group. Thus, there was a 50% rate of dyslexia among those children who had one older brother or sister with dyslexia (genetic risk) and approximately a 62% rate of dyslexia among those who had manifested an early language delay. These rates are in line with those from other previous studies of both genetic (Gilger, Pennington, & DeFries, 1991; Pennington, 1991; Snowling et al., 2003) and language delay (Aram & Hall, 1989; Bashir & Scavuzzo, 1992) risk. When we compared those 13 children each in the genetic and language delayed groups, respectively, who had dyslexia on all tasks administered at times 1 and 2, we found no differences in performances on any cognitive or lit- eracy tasks across time between them. Thus, for all subsequent analyses, these two groups were com- bined to form a larger group with dyslexia. All chil- dren who had been at risk for dyslexia but did not manifest the disorder at time 2 were then labeled the non-dyslexic group.
About eight months following time 2 testing, we telephoned parents of as many at-risk children as possible to determine whether any of them were receiving any special interventions at school. Of the 21 out of 26 dyslexic children of families who could be contacted (4 families were unreachable after repeated calls), 10 were currently receiving some therapy at school, and 1 additional child had had
some therapy previously but was no longer receiving it. Of the 19 non-dyslexic children whose families were reachable by phone, 7 of them had previously experienced some type of therapy at school; of these 7, 4 were currently still receiving it. Most of the intervention therapies were focused on articulation and speech organization; a minority incorporated orthographic knowledge training as well. Beyond these language-related therapies, some of the chil- dren additionally received occupational therapy, focused more specifically on fine motor skills and attention. Each therapy session at school lasted from 15 minutes to 1.5 hours; some occurred monthly, and others occurred weekly.
Our focus shifted next to a comparison of the performances among the three groups (i.e., dyslexic, non-dyslexic, and controls) across the six metalin- guistic tasks administered two years previously. The three groups differed in terms of age at time 2 (with the mean ages for at-risk dyslexic, at-risk non-dys- lexic, and control children being 88.8, 85.6 and 90.9 months, respectively, F(2,91) = 5.65, p < .05). The children who were at risk for dyslexia but non-dys- lexic differed in age from the other two groups, which did not differ from one another. However, the groups differed neither in mothers’ education level (F(2,91) = .28, n.s.) nor in nonverbal intelligence (F(2,91) = .47, n.s.) at time 1. Therefore, the task performances were compared using ANCOVA, with age at time 2 as the only covariate. Results of ANCOVAs (Table 1) showed that there was a significant overall group difference across all experimental tasks except for Visual-Spatial Relationships, syllable deletion, and tone detection.
The performances of the control children were significantly better than those in both at-risk groups (dyslexic and not) in rapid number naming, F(2, 90) = 5.52, p < .01, partial g2 = .11. In contrast, for the morphological awareness tasks, the control group scored significantly higher than those 26 children with dyslexia only, F(2, 91) = 3.73, p < .05, partial g2 = .08. For Chinese character recognition at time 1, the at-risk non-dyslexic performed as poorly as the at-risk dyslexic group and worse than the control group (F(2, 91) = 12.73, p < .001, partial g2 = .22).
Apart from looking at group differences, we per- formed correlational and regression analyses to explore the role of the metalinguistic measures as early predictors of children’s subsequent reading and writing outcomes. Because there was some evidence that the at-risk non-dyslexic group scored between the at-risk dyslexic and control groups (i.e., significantly higher than the dyslexics but lower than the controls) and also for simplicity, correlations are shown in Table 2 separately only for the at-risk and control groups, though the patterns of correlations for the dyslexic and non-dyslexic at-risk groups were similar. Because mothers’ education levels were not significantly associated with any of the perfor- mance measures, this variable was not included in
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Early predictors of dyslexia in Chinese children 207
208
Catherine McBride-Chang et al.
correlational or any subsequent analyses. However, because age was significantly associated with some variables, partial correlations among all variables are shown in Table 2 with age statistically con- trolled.
We then examined the extent to which we could predict individual literacy skills two years later, based on time 1 cognitive tasks, in the entire group com- bined in order to maximize statistical power, given the relatively large number of variables included. We included sex in these analyses because for dictation, but not for the other variables, girls significantly outperformed boys. Given inter-correlations among predictors, we report sr2 (in addition to b) in Table 3 to reflect the unique variance explained by each pre- dictor after accounting for the variances contributed by the other predictors combined.
Without controlling for Chinese word reading at time 1, between 26% and 31% of the variance could be explained by these variables across groups at time 2. The strongest correlate of word reading was mor- phological awareness; morphological awareness was also marginally (p < .07) significantly associated with one-minute reading and dictation. Speeded naming was a unique predictor of one-minute reading and marginally associated with dictation. Age tended to be significantly negatively associated with these literacy skills because we used scaled, rather than raw, scores to predict these variables. (Because some of these assessments were made by busy clinicians, raw scores were not always available for these analyses.) Age was already taken into account in the scaled scores, but it was included here because of its positive associations with some of the cognitive variables (i.e., morphological awareness, syllable deletion, and Visual-Spatial Relationships) at time 1. Overall, these analyses indicate that Hong Kong Chinese children’s skills in speeded naming and morphological aware- ness may be among the most sensitive predictors of subsequent performance on literacy skills. Syllable deletion, tone detection, and visual skills, though correlated with subsequent literacy skills, were not unique predictors of subsequent Chinese word reading in this combined group.
We also carried out these analyses for the three literacy skills with the autoregressive effects of Chi- nese word reading at time 1 included. In these anal- yses, morphological awareness remained a unique predictor of subsequent word reading only. In con- trast, only speeded naming was a unique predictor of subsequent speeded reading. None of the cognitive tasks were unique predictors of dictation skill apart from early word reading itself. These analyses indi- cate that even for very young Hong Kong Chinese children, Chinese word reading itself, if available, is the best predictor of subsequent literacy skills.
Another way to analyze these data is to determine which variables uniquely explain variance in the lit- eracy tasks using stepwise (forward) regression analyses. We did this separately for the control and
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Table 2 Partial correlations among time 1 metalinguistic measures and time 2 literacy outcomes (word reading, one-minute reading and word dictation) with time 1 age controlled Control (N = 47)
At-risk (N = 47) Nonverbal intelligence
Visual-spatial relationships Rapid number naming Morphological awareness Syllable deletion
Tone detection Chinese character recognition Chinese word reading (T2) One-minute reading (T2) Chinese word dictation (T2)
Rapid Nonverbal Visual-spatial number Morphological
intelligence relationships naming awareness .22 .12 .07
.61 ).01 .11 ).08 ).46 ).02
.39 .31 ).34 .13 .35 ).48 .28 .36 .30 ).16 .33 .16 .16 ).21 .55 .16 .33 ).14 .33 .13 .33 ).39 .34 .16 .34 ).18 .22
Chinese Chinese Syllable Tone character word
deletion detection recognition reading (T2)
.13 .18 .11 .12
.31 .38 .21 .05 ).27 ).27 ).22 ).18 .28 .57 .25 .34 .40 .26 .16 .28 .56 .42 .37 .20 .55
.19 .08 .47 .29 .04 .32 .67 .08 ).01 .22 .72
One-minute reading (T2)
.18 ).02 ).49 .12 .09 .37 .45 .57
.56
Chinese word dictation (T2)
.15
.05 ).32 .19 .43 .28 .40 .57 .50
Note: Correlations of magnitude.31 or higher were significant at p < .05.
Early predictors of dyslexia in Chinese children 209 Table 3 Final standardized beta weights explaining three literacy outcomes at time 2 from metalinguistic measures at time 1
Age ).38 Sex .05
Nonverbal intelligence Visual-spatial relationships Rapid number naming Morphological awareness Syllable deletion
Tone detection Chinese word reading R2 (df)
).23 .03 .06 ).06 ).39 .07 ).06 ).03 .48 .46
).20* ).22 .03 .06 .05 .03
).05 ).04 ).33*** ).43 .06 .21 ).05 ).00 ).03 .09
.40*** (9, 82) .31
).19* .06 .16 .03 .04
).26** ).29 .15+ .18
).25** .17+
With T1 read- ing
b sr2
Word reading
One-minute reading
Dictation
Without T1 reading
b sr2
).32** .07
).05 .07 ).11
.30** .02 .05
.55 .45*** .48 (9, 82) .28 (8, 84)
With T1 read- ing
b sr2
Without T1 reading
b sr2
With T1 read- ing
b sr2
Without T1 reading
b sr2
).33*** ).37 .04 .07 ).02 ).02 ).06 .08 .06 .09 ).07 ).06 ).13 .21 .17* .36 ).04 ).04 .02 ).07 ).06 .07
).30 ).03 .01
+p < .10, *p < .05, **p < .01, ***p < .001. sr2 represents the part correlation, i.e., the correlation between the predictor and the criterion variable with other predictors statistically controlled.
at-risk groups. (Because the results for the whole combined sample mirrored those shown in Table 3, they are not reviewed here.) For the control group, with word reading from time 1 included, only time 1 word reading uniquely emerged as a correlate of word reading at time 2 (total R-squared = .26), only speeded naming and time 1 word reading emerged as predictors of One-Minute Reading (total R-squared = .33), and only syllable deletion emerged as a unique correlate of dictation (total R-squared = .16). In contrast, the unique predictors for the at-risk chil- dren were age, word reading at time 1, and Visual- Spatial Relationships for word reading at time 2 (total R-squared = .45), age and speeded naming for One-Minute Reading (total R-squared = .20), and age and Visual-Spatial Relationships for dictation (total R-squared = .24). Patterns remained the same when word reading at time 1 was excluded in the analyses for the control and at-risk groups separately, with one exception: only tone detection was a unique predictor of word reading in the control group (total R-squared = .16).
Discussion
The present study was a follow-up on previous work (McBride-Chang et al., 2008) seeking to identify those Hong Kong Chinese children at greatest risk for dyslexia at a very young age based on a battery of cognitive tasks. We demonstrated that 50% of those at genetic risk for reading difficulties and 62% of those with a diagnosed language delay at age 5 subsequently manifested dyslexia in this sample. The genetic and language delayed groups did not differ from one another in severity of dyslexia. When groups of children who had and had not subse- quently manifested dyslexia at age 7 were compared to one another and to a control group, neither the visual spatial skill nor the lexical tone or syllable
deletion tasks were clear unique predictors of sub- sequent reading difficulties. However, measures of speeded naming and morphological awareness did distinguish the children who were ultimately diag- nosed as dyslexic from a control group. Moreover, only these two measures emerged as unique pre- dictors in the entire group of one or more subsequent literacy skills. In addition, very early word reading was the best measure for distinguishing among children who subsequently manifest the diagnosis of dyslexia.
These results are practically important. They underscore the overlap between language delay and dyslexia found in previous work in Hong Kong (Wong et al., 2010). In addition, they highlight the impor- tance of going beyond phonological sensitivity mea- sures only to diagnose subsequent reading difficulties in Chinese children. As demonstrated in correlational studies of older Chinese children with dyslexia, both speeded naming (Ho & Lai, 1999) and morphological awareness (Chung et al., 2008; Shu et al., 2006; Wong et al., 2010) are key predictors of reading difficulties in Chinese children. Moreover, given the unique educational structure of Hong Kong, a city in which formal literacy instruction begins very early (e.g., Cheung & Ng, 2003), it is equally vital to note that Chinese word reading itself at age 5 tends to be the best predictor of subsequent reading difficulties in older children. At the same time, however, reading itself is strongly influenced by explicit teaching. Therefore, clinicians should use their own judgment in assessing the importance or weight assigned to early word reading in predicting subsequent reading difficulties in at-risk groups.
Theoretically, these results also demonstrate the broad array of metalinguistic and cognitive skills that are potentially important for learning to read Chinese. Although phonological sensitivity is clearly developmentally vital for learning to read Chinese, the unreliability of phonological cues to Chinese
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
).37*** .17+
).00 .07
(8, 84)
.03
.01 ).17 ).14 .11 .08 .08 .07 ).09 ).07
.44 .37*** .39 (9, 82)
.01 .01
.02 .02 ).21 ).18+ .23 .19+
.14 .11 .02 .02
.26 (8, 84)
210 Catherine McBride-Chang et al.
words, in addition to the relatively easy phonological structure of the Chinese language, makes it essential to look at other indicators of reading difficulties. Morphological awareness may be particularly helpful in early acquisition of Chinese literacy skill. In addition, fluency in identifying printed symbols, in this case in the form of simple numbers, is a strong correlate of subsequent reading, particularly spee- ded reading.
At the same time, forward stepwise regression analyses by group also indicated that all of the measures included at time 1 were uniquely asso- ciated with at least one time 2 literacy skill in either the control or at-risk group. For example, both the syllable awareness and tone detection measures explained unique variance in subsequent literacy skill in the control group. Thus, within this group, with, almost by definition, greater variability across all literacy tasks at time 2 as compared to those who were initially at risk for reading difficulties, phonological sensitivity measures were significant predictors of subsequent reading and writing. In contrast, in the at-risk group, early visual, but not phonological sensitivity, skills were uniquely asso- ciated with subsequent dictation and untimed word reading, suggesting that visual skills may be useful, perhaps somewhat independently of (phonological or morphological) metalinguistic skills, for Chinese literacy acquisition among at-risk readers. Future work should consider the possibility that the route to Chinese literacy acquisition may include multiple dimensions which may differ somewhat across those with and without risks for reading difficulties.
Importantly, those at-risk children who ultimately did not manifest dyslexia two years later neverthe- less performed significantly more poorly than the control group on two of the three literacy skills tested at time 2, though they had differed from the control group at time 1 only on speeded naming and word reading. This pattern is similar to one demonstrated in a longitudinal study of English-speaking children at familial risk for dyslexia and adds weight to the argument (Snowling et al., 2003) that solid language skills may help children at risk for dyslexia to com- pensate for some reading difficulties. At the same
time, the fact that this at-risk group was initially slower on the rapid naming task also suggests that automatization of literacy skills may be a continuing challenge for them.
The present study was limited in the measures it included. As noted previously (McBride-Chang et al., 2008), ours was not an exhaustive group of cognitive tasks. For example, paired associate learning might have been a good measure of early learning to include (Li, Shu, McBride-Chang, Liu, & Xue, 2009). Future studies might increase the number of mea- sures tested to predict subsequent reading difficul- ties. In addition, there were some age-diagnosis interactions. Indeed, the children who had been initially at risk for dyslexia but did not manifest the disorder were significantly younger than were those of the same grade level who were diagnosed with dyslexia. Criteria for the diagnosis of dyslexia are stricter for older children.
Despite these limitations, however, this study has been important in at least two ways. First, we have examined longitudinally in a prospective study the overlap between genetic risk, as well as early lan- guage delay, and subsequent dyslexia in Chinese. This overlap is considerable and adds to the growing understanding of how important early screening is for later reading problems. Second, we have isolated speeded naming and morphological awareness, apart from reading itself, as two cognitive tasks that appear to be potentially clinically useful in identify- ing very young Chinese children at risk for sub- sequent reading difficulties.
Acknowledgements
We are grateful to the Research Grants Council of the Hong Kong Special Administrative Region (project reference #448907) for support of this research.
Correspondence to
Catherine McBride-Chang, Psychology Department, The Chinese University of Hong Kong, Shatin, Hong Kong, China; Tel: 852-26096576; Email: cmcbride@ psy.cuhk.edu.hk
Key points
• The rate of dyslexia at age 7 in Chinese children with a language delay at age 5 was 62% and with a genetic risk for dyslexia at age 5 was 50%.
• The cognitive tasks at age 5 that best distinguished Chinese children with and without dyslexia at age 7 and that predicted subsequent literacy skills across all children were rapid automatized naming and morphological awareness. At the same time, word reading at time 1 itself was the best predictor of sub- sequent reading difficulties in this sample.
• Within the control group only, syllable deletion and tone sensitivity were significantly associated with subsequent literacy skills; among the at-risk group, visual skills were significantly associated with sub- sequent literacy skills.
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
References
Aram, D.M., & Hall, N.C. (1989). Longitudinal follow-up of children with preschool communication disorders: Treatment implications. School Psychology Review, 18, 487–501.
Bashir, A.S., & Scavuzzo, A. (1992). Children with lan- guage disorders: Natural history and academic success. Journal of Learning Disabilities, 25, 53–65.
Chan, D.W., Ho, C.S.-.H, Tsang, S.-M., Lee, S.-H., & Chung, K.K.H. (2006). Exploring the reading-writing connection in Chinese children with dyslexia in Hong Kong. Reading and Writing: An Interdisciplinary Journal, 19, 543–561.
Cheung, H., & Ng, L. (2003). Chinese reading development in some major Chinese societies: An introduction. In C. McBride-Chang, & H.-C. Chen (Eds.), Reading develop- ment in Chinese children (pp. 3–17). Westport, CT: Praeger Press.
Chung, K.K.H., McBride-Chang, C., Wong, S.W.L., Che- ung, H., Penney, T.B., & Ho, C.S.-H. (2008). The role of visual and auditory temporal processing for Chinese children with developmental dyslexia. Annals of Dys- lexia, 58, 15–35.
Ciocca, V., & Lui, J.Y.K. (2003). The development of the perception of Cantonese lexical tones. Journal of Multi- lingual Communication Disorders, 1, 141–147.
Gardner, M.F. (1996). Test of visual-perceptual skills (non- motor). Burlingame, CA: Psychological and Educational Publications.
Gilger, J.W., Pennington, B.F., & DeFries, J.C. (1991). Risk for reading disability as a function of parental history in three family studies. Reading and Writing: An Interdisci- plinary Journal, 3, 299–313.
Ho, C.S.-H., Chan, D.W., Chung, K.K.H., Lee, S.-H., & Tsang, S.-M. (2007). In search of subtypes of Chinese developmental dyslexia. Journal of Experimental Child Psychology, 97, 61–83.
Ho, C.S.-H., Chan, D.W.-O., Lee, S.-H., Tsang, S.-M., & Luan, H.V. (2004). Cognitive profiling and preliminary subtyping in Chinese developmental dyslexia. Cognition, 91, 43–75.
Ho, C.S.-H., Chan, D.W.-O., Tsang, S.-M., & Lee, S.-H. (2000). The Hong Kong Test of Specific Learning Difficul- ties in Reading and Writing. Hong Kong: Hong Kong Specific Learning Difficulties Research Team.
Ho, C.S.-H., & Lai, D.N.-C. (1999). Naming-speed deficits and phonological memory deficits in Chinese develop- mental dyslexia. Learning and Individual Differences, 11, 173–186.
Hulme, C., & Snowling, M.J. (2009). Developmental disor- ders of language learning and cognition. Chichester: Wiley-Blackwell.
Li, H., Shu, H., McBride-Chang, C., Liu, H.-Y., & Xue, J. (2009). Paired associate learning in Chinese children with dyslexia. Journal of Experimental Child Psychology, 103, 135–151.
McBride-Chang, C., Cho, J.-R., Liu, H., Wagner, R.K., Shu, H., Zhou, A., Cheuk, C.S.-M., & Muse, A. (2005). Changing models across cultures: Associations of pho- nological and morphological awareness to reading in Beijing, Hong Kong, Korea, and America. Journal of Experimental Child Psychology, 92, 140–160.
McBride-Chang, C., Lam, F., Lam, C., Doo, S., Wong, S.W.L., & Chow, Y.Y.Y. (2008). Word recognition and cognitive profiles of Chinese pre-school children at-risk for dyslexia through language delay or familial history of dyslexia. Journal of Child Psychology and Psychiatry, 49, 211–218.
Pennington, B.F. (1991). Diagnosing learning disorders: A neuropsychological framework. New York: Guilford Press.
Raven, J.C. (1956). Coloured progressive matrices. Oxford: Psychologists Press.
Shaywitz, S.E. (2003). Overcoming dyslexia. New York: Alfred A. Knopf.
Shu, H., McBride-Chang, C., Wu, H., & Liu, H. (2006). Understanding Chinese developmental dyslexia: Mor- phological awareness as a core cognitive construct. Journal of Educational Psychology, 98, 122–133.
Snowling, M.J., Gallagher, A., & Frith, U. (2003). Family risk of dyslexia is continuous: Individual differences in the precursors of reading skill. Child Development, 74, 358–373.
Wong, A.M.Y., Kidd, J., Ho, C.S.-H., & Au, T.K.-F. (2010). Characterizing the overlap between SLI and dyslexia in Chinese: The role of phonology and beyond. Scientific Studies of Reading, 14, 30–57.
Manuscript accepted 24 May 2010
Ó 2010 The Authors Journal of Child Psychology and Psychiatry Ó 2010 Association for Child and Adolescent Mental Health.
Early predictors of dyslexia in Chinese children 211
It is a great concern that young children in China will be Dysexia
Subscribe to:
Posts (Atom)