READERS with dyslexia have pronounced

Top Lang Disorders Vol. 37, No. 2, pp. 170 181 Copyright c 2017 Wolters Kluwer Health, Inc. All rights reserved. English Word Reading Difficulties and Orthographic Processing Weaknesses in Chinese English Bilingual Adolescents With Dyslexia Xiuhong Tong and Catherine McBride Is dyslexia in Chinese for Chinese English bilinguals associated with difficulties in reading English, given differences in L1 and L2 orthographies? Among 11 Hong Kong Chinese adolescents with dyslexia, who were diagnosed by professional psychologists using the diagnostic criteria set out in a standardized test, and 14 adolescents without dyslexia, Chinese word reading was tested at the age of 9 years; English word reading was tested across ages 9, 10, 11, and 12 years; and English orthographic processing was tested in a lexical decision task at the age of 13 years. The lexical decision task required participants to judge whether or not stimuli appeared to look like a possible real word in English across 3 conditions (real words, look-like words that were orthographically possible in English, and nonwords that violated orthographic rules of English). English word reading differed significantly between the 2 groups in 2 of the 4 years. Both groups found it easier to identify nonwords that violated English orthographic rules than those that did not. However, compared with peers without dyslexia, adolescents with dyslexia had more difficulties with English orthographic rules in accuracy but not reaction time, suggesting that children with dyslexia may manifest specific difficulties in English orthographic processing. Key words: biliteracy development, Chinese dyslexia, English orthographic processing Author Affiliations: Institute of Psychological Sciences, Zhejiang Key Laboratory for Research in Assessment of Cognitive Impairments, and Center for Cognition and Brain Disorders, Hangzhou Normal University, Hangzhou, Zhejiang Sheng, China (Dr Tong); and Psychology Department, Hong Kong Special Administrative Region, The Chinese University of Hong Kong, Shatin, NT (Dr McBride). This research was supported by the Collaborative Research Fund (CUHK CUHK8/CRF/13G) and a Grant of the BRAIN initiative of the Lui Che Woo Institute of Innovative Medicine, The Chinese University of Hong Kong (Project Code: 8303105), to Catherine McBride. We thank all those who helped in the data collection and children and parents for their participation. The authors have indicated that they have no financial and no nonfinancial relationships to disclose. Corresponding Author: Catherine McBride, PhD, Psychology Department, The Chinese University of Hong Kong, Shatin, Hong Kong (cmcbride@psy.cuhk.edu.hk). DOI: 10.1097/TLD.0000000000000119 170 READERS with dyslexia have pronounced difficulty with word reading and spelling in their first language (L1), despite normal intelligence, adequate schooling, and reasonable social economic opportunities (Lyon, Shaywitz, & Shaywitz, 2003). The extent to which students with dyslexia have difficulties with reading in a second language has been examined relatively rarely so far, particularly among Chinese students learning English as a second orthography. This is of particular interest given the differences in orthographies between Chinese and English languages. In this study, we examined English word reading difficulties in Chinese children who had been diagnosed with dyslexia in their first language (usually Cantonese). We further tested these children s orthographic pattern recognition skills in English and compared their

English Word Reading Difficulties and Orthographic Processing Weaknesses 171 skills with those of Chinese children without dyslexia when reading English. It is a challenge for bilingual children to be proficient in both L1 and L2 reading. This is particularly evident among Chinese English bilingual children who learn two distinct writing systems. Chinese and English differ fundamentally in orthography (Geva & Siegel, 2000), which is defined here as the rules for writing a language, such as conventions of spelling and grapheme phoneme correspondences. Chinese is a morphosyllabic language in which a character represents a syllable. Most characters (>80%) are semantic phonetic compounds that are composed of semantic and phonetic radicals. Unlike the sublexical orthographic units or letter clusters in English, many radicals exhibit orthographic, phonological, and semantic properties. In addition, radicals comprising a character have positional regularities. For example, the Chinese character (, stare) is composed of the semantic radical (eye) and the phonetic radical ( ); the semantic radical (eye) is usually located ontheleftsideandthephoneticradical ( ) often appears on the right side in a left right structured character. Such unique features of Chinese orthography instantiate the concept of orthography semantics (i.e., semantic radicals) and orthography phonology semantics (i.e., phonetic radicals) processing both of the pathways are present in the writing itself (Seidenberg, 2011, p. 162). In contrast, English has different visual forms from Chinese. That is, letters or letter clusters are arranged linearly in English. The mapping principle between graphic units and sounds in English is also different from Chinese (e.g., Perfetti et al., 2007; Tong, Tong, & McBride-Chang, 2015). Although the mappings are not simple oneto-one correspondences between letters and sounds in English, English learners rely more on the phonological mediation route of orthography phonology semantics than the direct route from orthography to semantics in word recognition. Given the basic differences between Chinese and English in orthography, it was of both theoretical and practical interest to examine whether Chinese children with dyslexia in their L1 would show weakness in L2 English word reading and whether those children would have problems in English orthographic processing. Models of second language learning suggest that L1 and L2 literacy skills are interrelated. More specifically, there are some common cognitive and linguistic processes that underlie both L1 and L2 reading such that the learning of L2 is similar to that of learning L1 (Cummins, 1984; Koda, 2005). For example, the linguistic interdependence theory holds the view that reading abilities in L1 and L2 are interdependent and certain cognitive and linguistic skills acquired in L1 learning can be positively transferred to L2 acquisition; this transfer is helpful to the development of corresponding abilities in L2 (Cummins, 1979). A number of empirical studies have reported modest correlations between L1 and L2 reading performance (e.g., Bernhardt & Kamil, 1995; Van Gelderen et al., 2004), which, in part, supports the comorbidity hypothesis between L1 and L2 reading. It is thus theoretically possible, and perhaps even probable, that readers with dyslexia in L1 may have a difficulty in L2 reading (Tong et al., 2015). Several factors might account for the L2 reading difficulties. For example, deficits in L1 phonological processing skills might be one of the factors that correlate with reading difficulties in L2 and its influence on L2 might be the same as it is on L1 reading (Wade-Woolley & Geva, 2000). Indeed, Tong and colleagues reported that Chinese English bilingual readers who were poor only in L2 English scored significantly lower in L1 Chinese tone awareness, a suprasegmental phonological processing skill in Chinese, in comparison with the adequate bilingual readers (Tong et al., 2015). However, there is relatively little research on the co-occurrence and dissociation of L1 Chinese and L2 English reading difficulty in dyslexic readers (Tong et al., 2015), and findings on this issue have been somewhat inconsistent (Zhou et al., 2014). Ho and Fong (2005)

172 TOPICS IN LANGUAGE DISORDERS/APRIL JUNE 2017 observed a high overlap of poor reading in English among those with dyslexia in Chinese. In particular, they found that among 25 dyslexic children with mean ages of 9 years, all except one had lower scores in both English word reading and English reading comprehension compared with control children. This finding was supported in another study, in which Chung and Ho (2009) reported a relatively high co-occurrence of L1 Chinese and L2 English reading difficulty in Chinese children with mean ages of 9.9 years. However, Zhou et al. (2014) did not find significant differences in English word reading between Chinese children with dyslexia and agematched controls at ages of 6, 7, and 8 years. In this study, we tested for difficulties in English word reading among the general population of Chinese children in Hong Kong and among those diagnosed as having dyslexia in Chinese. Children learn Chinese and English in parallel in Hong Kong, although Cantonese is usually the medium of classroom instruction. Hong Kong children start to learn English in schools at around the age of 3.5 years. It was thus of practical importance to examine the question of whether children with dyslexia in their L1 would show difficulties in their L2 English word reading as well. In addition, if those Chinese children with dyslexia had a problem in English word reading, we wondered whether they might show special weakness in English orthographic processing. This was the second research question examined in this study. The theoretical dual route model of L1 reading suggests that there are two routes for English word reading, namely, the sublexical phonological route and the lexical orthographic route, with the two routes interacting in word reading at some points in some contexts (Coltheart, Rastle, Perry, Langdon, & Ziegler, 2001). Numerous studies in readers of alphabetic languages have suggested that most of those native English speakers who demonstrate dyslexia have difficulties in using the sublexical phonological route (i.e., they have deficits in grapheme phoneme conversion) (Snowling, 1998). In addition, orthographic processing, which may be defined as the ability to form, store, and access orthographic representations (Stanovich & West, 1989; p. 404), is clearly an indicator of reading ability in native English speakers (for a review, see Burt, 2006). That is, readers with reading difficulties show poor performance in orthographic processing compared with typically developing children (Corcos & Willows, 1989; Hanley, Hastie, & Kay, 1992). We wondered whether Chinese children with dyslexia, who were learning to read English, would manifest difficulty in English orthographic processing. English orthographic processing can be measured using a lexical decision task, in which participants are required to make a judgment as to whether a word presented is likely to be a real word or not (Burt, 2006). In the present study, we modified this lexical decision task by asking participants to judge whether words appeared to them to look like words or not. This task was similar to the orthographic constraints test designed by Treiman (1993), which has been used to tap children s orthographic knowledge (Cassar & Treiman, 1997). We recorded participants reaction time and response accuracy rate to three types of stimuli, consisting of (1) real English words (e.g., cat), (2) look-like words (i.e., words that were orthographically possible in English, but do not exist in English e.g., lafe), and (3) nonwords that did not look like possible English words (i.e., were orthographically impossible in English e.g., rmdu). In summary, there were two research objectives of this study. The first goal was to investigate whether students with dyslexia in Chinese would manifest difficulties in English word reading at ages of 9 12 years. We expected to find more difficulties among Chinese children with dyslexia in English word reading than among the control group. It was also possible that those children with L1 Chinese word reading difficulty might not have a problem in L2 English word reading at certain ages. Our second objective was to tap Chinese English bilingual adolescents

English Word Reading Difficulties and Orthographic Processing Weaknesses 173 sensitivity to orthographic patterns at the age of 13 years. Here, we expected that participants would have more difficulty in making a judgment about look-like words than nonwords because even participants with dyslexia would be able to notice orthographic violations in English and reject nonwords quickly. However, those with dyslexia were expected to show longer reaction time or lower accuracy rate in judging all three types of stimuli as compared with those without dyslexia. METHODS Participants In total, 11 students with dyslexia (5 females and 6 males) with a mean age of 13.36 years (SD = 0.50 years) and 14 typical developing readers (10 females and 3 males) with a mean age of 13.71 years (SD = 0.47 years) participated in this study. These participants had participated in a 10-year longitudinal study since their early childhood. Thus, they had complete data at all measurement points included in this study. Originally, there had been 298 children involved in this 10-year longitudinal study, which focused on Chinese children s L1 and L2 language acquisition and literacy development. These children were recruited from Maternal and Child Health centers from three districts of Hong Kong, spanning across the New Territories, Kowloon, and Hong Kong Island, and they were drawn from comparable backgrounds in terms of the family s social economic status and parents education. Among those 298 children, there were a few who were officially diagnosed as having developmental dyslexia in the first year of their primary school. Those 11 children with dyslexia and 14 children in the control group were involved in the longitudinal study. All children lived in Hong Kong and were native Cantonese speakers. None of the participants had a visual or hearing problem or a history of other learning difficulties per parents reports. The dyslexic children were diagnosed by professional psychologists using the diagnostic criteria set out in the Hong Kong Test of Specific Learning Difficulties in Reading and Writing (HKT-P (II); Ho et al., 2007). According to HKT-P (II), children were assessed using four subtests of cognitive linguistic skills (i.e., speeded naming assessed by digital rapid naming, phonological awareness assessed by rhyme detection and onset detection, phonological memory assessed by word repetition I and II and nonword repetition, and orthographic pattern recognition assessed by left-right reversal, lexical decision and radical position tasks) and literacy skills (i.e., word reading, 1-min word reading, and dictation). However, nonphonological language skills associated with sentence and discourse level reading were not included in the assessment battery. Children s visual perceptual and visual memory skills also were assessed by using the Test of Visual-Perceptual Skills (Non-motor) Revised (TVPS-R). Children who were classified as having dyslexia were those whose literacy composite score and one cognitive-linguistic composite score were at least 1 standard deviation below their respective age means according to the HKT-P (II) and TVPS-R. The control children were randomly selected from the remaining 149 children who manifested typical reading skills. In addition, according to the policy of the Hong Kong Education Department (1996), Hong Kong children begin receiving formal instruction in reading and writing in Chinese at around the age of 3.5 years. Children begin learning Chinese compound words and short phrases in K2 (second year of kindergarten, around the age of 4 years). Moreover, children typically begin to learn English at school at the age of 3.5 years or even earlier (McBride- Chang et al., 2008), and they are formally taught English in kindergarten, either by local English teachers whose native language is Chinese, or by native English speakers. In addition, some Hong Kong families hire Filipina women as domestic helpers to look after their children. The Filipinas often speak English to the children. Hong Kong parents are very

174 TOPICS IN LANGUAGE DISORDERS/APRIL JUNE 2017 motivated to speak with their children in English. Although we did not formally assess children s English proficiency, Hong Kong Chinese children by the age of 13 years typically have had at least 9 years of exposure to English. Measures This was a retrospective study based on existing data gathered as part of a 10-year longitudinal study. The measures, as described later, were administered at different ages, spanning from 4 to 13 years. For example, nonverbal intelligence (IQ) was tested at the age of 4 years. The data for Chinese word reading included in this study were from age 9 years. English word reading data were collected at ages of 9, 10, 11, and 12 years. The English orthographic processing task was administered when children were 13 years old. We focused on L1 word-level learning in the early years of development, and we also collected several additional reading-related measures each year. Our focus was primarily on L2 orthographic skills and L1 and L2 reading comprehension and writing skills in the later years. We did not have the resources to collect a broader battery of measures than this. In this study, we report only on word-recognition skills at ages 9, 10, 11, and 12 years, and performance on the English orthographic processing measure at the age of 13 years. Nonverbal reasoning This task was measured one time, at the age of 4 years. Raven s Progressive Matrices was used to test children s nonverbal reasoning ability (Raven, Court, & Raven, 1995). There were 24 colorful pictures with a portion missing. Children were asked to select the item that best fit the matrix from among six choices. One point was given for each correct response. Therefore, the maximum raw score was 24 for this task. English orthographic processing task English orthographic processing was measured once, at the age of 13 years. This task was designed to tap children s English orthographic processing using a modified lexical decision paradigm. In this task, there were three experimental conditions, including real word, look-like word (i.e., pseudowords), and nonword conditions with 50 test items for look-like and nonword conditions and 100 items for the real word condition. Children were asked to judge whether stimuli appeared to resemble a possible English word. Thus, both real English words and look-like words should have been accepted whereas the nonwords should have been rejected in this study. Real words were selected from children s English textbooks used in kindergarten and primary schools in Hong Kong. Those that were not word-like contained a systematic violation of orthographic principles of English (e.g., kcdu). Those stimuli that were word-like (e.g., pait) were not real words but followed the rules of English orthography. All stimuli were four letters long. The look-like words and nonwords were created on the basis of the same four-letter real words. For example, the stimulus of lafe in the look-like word condition was developed by replacing the letter i in the real word life with letter a. The nonwords were created by rearranging the orders of the four letters comprising real words. For example, the nonword hspo was created by randomly arranging the four letters in the real word shop. A native-english speaker was then invited to proof whether the stimuli in look-like word condition and nonword condition were clearly pseudowords and nonwords. In addition, consonant doublets at beginning of words and vowel doublets were not allowed in any of the three conditions. E-Prime software (E-Prime 2.0, Psychology Software Tools, Inc., Pittsburgh, PA, USA) was used to design the testing program and record children s reaction times and accuracy rates. All stimuli were presented visually in a white against black background at the center of the computer screen. A fixation + wasfirst presented at the center of a computer screen for 500 ms followed by a blank with a mean of 500 ms. The target stimulus was then presented

English Word Reading Difficulties and Orthographic Processing Weaknesses 175 for a total of 400 ms. There was a 1000-ms interval after the presentation of each stimulus. The stimuli were presented in a random order. The participants were required to press a button on the keyboard as quickly and as accurately as possible. To make sure that they understood the task, students were not allowed to continue the experiment until their accuracy rate reached at least 50% in the practice section. English word reading English word reading was measured annually at ages of 9, 10, 11, and 12 years. Children at these ages were asked to read aloud 60 English words that were selected from English textbooks used in Hong Kong kindergarten and primary schools. This task was developed on the basis of a task used in a previous study (Tong & McBride-Chang, 2010). One point was given for each correct word, and the maximum possible score of the task was 60. One set of stimuli was used at the age of 9 years. This set was modified for children aged 10 12 years by replacing some of the items with more difficult words. The test stopped when children made five errors in a row. Chinese word reading Chinese word reading reported in this study was measured at the age of 9 years. It was measured using the Chinese word reading subtest of the Hong Kong Test of Specific Learning Difficulties in Reading and Writing (HKTSpLD; Ho, Chan, Tsang, & Lee, 2000). Children aged 9 years were asked to read aloud 150 compound words. One point was given for each correct compound word, and the maximum possible score of the task was 150. The test stopped when the children read 15 consecutive words incorrectly. Procedure The English orthographic processing task was tested individually in a laboratory at the university when the participants (i.e., 11 dyslexia and 14 controls) were aged 13 years. All other measures were conducted during the summer months (from June to September) at participants homes for each year. We recruited psychology majors to do the testing. Before the real testing, the testers were trained by professional research assistants who were familiar with each task used in each year. We sent invitation letters and consent forms to participants parents prior to the testing. After receiving signed consent forms, the testers contacted the families who agreed to participate and arranged a convenient testing time for the children and their caregiver. All tests for each year were administered in a single session. Supermarket coupons were given after the testing to express our gratitude for their participation. RESULTS Examining whether students with dyslexia had difficulty in English word reading Table 1 shows the means and standard deviations of all reading measures and dictation measures across the dyslexic and control groups. Student s t tests were performed to evaluate the differences between the groups in English word reading and Chinese word reading for each year. Results are shown in Figure 1. Not surprisingly, the group with dyslexia had a lower score in Chinese word reading at the age of 9 years compared with the control group, (t(22) = 6.44, p <.001), but the two groups did not show a difference in nonverbal reasoning abilities at the age of 4 years (t(22) = 1.33, p =.20). This confirms the diagnosis of dyslexia in that group. For English word reading, there was no significant difference between those with and without dyslexia (t(20) = 1.11, p =.28) at the age of 9 years. Although there was a significant difference between the two groups at ages of 10 and 11 years for English word reading (t(20) = 2.17, p <.05; t(19) = 2.86, p <.05, respectively), the group difference only approached marginal significance at the age of 12 years (t(20) = 1.63, p =.08).

176 TOPICS IN LANGUAGE DISORDERS/APRIL JUNE 2017 Table 1. Means and standard deviations (SD) for Chinese word reading at the age of 9 years and English word reading at ages of 9, 10, 11, and 12 years in two groups Dyslexic Group Control Group Measures M SD M SD t test Effect Size Nonverbal reasoning 11.00 2.21 12.21 2.19 1.33 0.54 Chinese word reading Age 9 37.27 14.13 98.31 28.57 6.44 *** 2.71 English word reading Age 9 10.82 9.77 15.54 11.16 1.11 0.45 Age 10 24.63 19.01 38.64 11.47 2.17* 0.89 Age 11 30.89 20.72 43.85 25.13 2.86* 0.56 Age 12 32.36 18.04 49.78 10.00 1.63 1.19 *p <.05; p =.08; *** p <.001. Evaluating whether students with dyslexia had difficulties in English orthographic processing Table 2 shows the means and standard deviations of reaction time and accuracy rates for all three conditions (real word, look-like word, and nonword) in the dyslexic and control groups. Repeated-measures analyses of variance (ANOVAs) with experimental conditions (real words vs. look-like word vs. nonword conditions) as a within-subject factor, and groups (dyslexic vs. typically developing groups) as the between-subject factor, were Figure 1. Comparisons of English word reading at ages of 9, 10, 11, and 12 years, and Chinese word reading at the age of 12 years between dyslexic and control groups. performed separately for reaction time and accuracy rates. Results are shown in Figure 2. The ANOVA for accuracy rate showed that there was a significant experimental condition effect, F(1, 23) = 13.47, p <.001, η p 2 = 0.37, with the lowest accuracy rate for looklike word condition (mean = 0.66) followed by the nonword condition (mean = 0.78) and the real word condition (mean = 0.83). The main effect of group was also significant, F(1, 23) = 8.94, p <.01, η p 2 = 0.28. Pairwise comparisons indicated that the dyslexic group had a lower accuracy rate than the control group overall (p <.01). The interaction between the group and the experimental condition was not significant, F(1, 23) = 0.44, p =.52. The ANOVA for reaction time revealed a significant main effect of experimental condition, F(1, 23) = 42.79, p <.001, η p 2 = 0.65. Pairwise comparisons showed that for both the dyslexic and control groups, the reaction times in both look-like word and nonword conditions were significantly longer than that for the real word condition, but there was no significant difference between the look-like word and nonword conditions. There was no main group effect, F(1, 23) = 0.96, p =.34, and the interaction of Group-by-Experimental condition was also not significant, F(1, 23) = 1.72, p =.20.

English Word Reading Difficulties and Orthographic Processing Weaknesses 177 Table 2. Means and standard deviations for accuracy and reaction time on the English orthographic processing task across the three conditions in the two groups at the age of 13 years Dyslexic Group (n = 11) Control Group (n = 14) Real Word Look-Like Word Nonword Real Word Look-Like Word Nonword F(1, 23) RT 640.86 (95.64) 663.17 (90.36) 689.83 (96.13) 591.72 (60.27) 650.28 (80.45) 665.29 (45.28) 8.94 ** Accuracy 0.73 (0.21) 0.60 (0.19) 0.70 (0.22) 0.92 (0.05) 0.73 (0.18) 0.86 (0.07) 0.96 Note. The numbers in parentheses are standard deviations. Accuracy is expressed as a proportion of possible correct responses. RT = reaction time. **p <.01. Figure 2. Reaction time, accuracy, and comparisons on the English orthographic processing task across conditions in the 2 groups at the age of 13 years. DISCUSSION Results of this study showed that word reading in English tended to be poorer among those who were dyslexic than for controls without dyslexia among Chinese children at ages 9 12 years. We also found that orthographic pattern recognition performance was poorer in accuracy for those with dyslexia as compared with those without dyslexia, although the groups were equally fast in making lexical decisions. In addition, for this orthographic processing task, both the dyslexic and control groups had more difficulty in making a judgment about look-like words compared with nonwords and real words, as reflected by accuracy rates. The finding that students with dyslexia in their L1 also showed difficulties in L2 word reading agrees with results observed in prior studies showing co-occurrence of L1 and L2 reading difficulties in Chinese English bilingual children (e.g., McBride-Chang et al., 2012; Tong et al., 2015). This finding is somewhat supported by the linguistic interdependence theory of L2 acquisition. This model assumes that only when L2 learners are fluent in their L1 reading, meaning that they have

178 TOPICS IN LANGUAGE DISORDERS/APRIL JUNE 2017 already developed strong language abilities to support the reading process, can they be fluent in their L2 reading. Conversely, L2 learners will be poor readers if they have difficulties in their L1 reading, meaning that they have not developed the language abilities to support the reading process. Alternatively, the same cognitive linguistic weaknesses that underlie problems in learning to read in L1 could have a similar influence on learning to read in L2. The most novel aspect of this study was that we designed an English orthographic processing task to test Chinese children s English orthographic processing. Our findings suggest that adolescents from both the dyslexic and control groups had developed a sensitivity to English orthographic knowledge as reflected by their lower accuracy rates to looklike words than to nonwords. Look-like words in our study were those stimuli orthographically similar to real words, and they were not easily categorized as nonwords. In contrast, nonwords were those with systematic violations of English orthography. It is highly possible that nonwords might have been rejected at first sight and without an orthographic search or semantic analysis being performed in making a judgment as to whether it was a real or look-like word. However, further orthographic and semantic analyses may be required for legal-formed look-like and real words; thus, it was more difficult to reject look-like words. These were more easily associated with making mistakes than were nonwords. The dyslexic group had an overall lower accuracy rate than the typically developing group in English orthographic processing, but the two groups did not differ in reaction time. This indicates that dyslexic children may not have difficulty in the speed of English orthographic processing (i.e., lexical retrieval), but the two groups seemed to have an overall accuracy level difference. That is, those with dyslexia may have had a weaker representation of English orthographic rules or they may have had smaller lexicons for mental orthographic image compared with those in the control group. These adolescents with dyslexia may have had a lower quality of lexical representation in English, which might have caused them to have difficulty visualizing letter strings or applying their knowledge of orthographic rules, such as using root words to decode a variation of a word, causing more errors in orthographic processing than control readers. This finding is consistent with the hypothesis that those with dyslexia might lag behind in lexical development and develop underspecified orthographic representations (Perfetti & Hart, 2002). Here, the focus is on this phenomenon in a second language. Clinical implications Our findings have at least two clinical implications. First, the finding of the co-occurrence of reading difficulty in L1 Chinese and L2 English suggests that children with reading difficulty in L1 might also have a problem in their L2 reading even when L1 and L2 have distinct linguistic and orthographic structures. This finding supports the hypothesis that many children with dyslexia in L1 have difficulties in their L2 reading. In a biliterate society such as Hong Kong, it is particularly challenging for children to be fluent in both L1 and L2 reading. While much attention has been given to children s L1 reading problems, the problems specific to reading difficulty in L2 English deserve attention as well and it appears that they have difficulties in both their L1 and L2. Early detection and intervention for bilingual children at risk for reading for both L1 and L2 should be important in a bilingual society. Alternatively, someone with difficulties in L1 reading could have fewer or no difficulties in L2 reading compared with peers (e.g., McBride-Chang et al., 2012; Tong et al., 2015). Indeed, findings in prior studies on poor readers in Chinese English bilingual children showed that there was dissociation between L1 and L2 reading difficulties (e.g., McBride-Chang et al., 2012; Tong et al., 2015). For example, Tong, Tong, and McBride-Chang found that the likelihood of those who were adequate readers of Chinese of being poor readers in English was 19% at Grade 2 and 15% at Grade 5. In

English Word Reading Difficulties and Orthographic Processing Weaknesses 179 the case of Chinese English bilinguals, we advocate, therefore, testing of reading separately in the L1 and L2 to target specifically those who need the most help in either script or in both. Second, the evidence of orthographic processing weakness for L2 English in adolescents with dyslexia suggests that orthographic processing difficulty in L2 is associated with the risk for reading difficulties in L2 for Chinese English bilingual children. This suggests that clinicians should consider incorporating English orthographic processing tasks into their standard evaluation batteries for L2 English reading difficulty in Chinese English readers, and those with reading difficulties in L2 English word reading should receive enhanced English orthographic training to strengthen their lexical representations of English words and to associate phonological and orthographic patterns in English. Study limitations There were some limitations of this study. The primary limitation was the small sample size of students with dyslexia, and that there were more females (i.e., 10 girls) in the control group than in the dyslexic group. Clearly results would be easier to interpret with a larger sample and balanced gender difference. This would enable us to see a clearer trend in English skills across the years. However, this was a longitudinal data set for which data were originally collected for a different purpose, and we did not select at all for dyslexia or familial history of dyslexia initially. Children were diagnosed with dyslexia during the course of the study. As we mentioned earlier, this study was a retrospective study, in which tasks were administered at different time points. In particular, neither L1 nor L2 word reading was measured when participants were 13. The English orthographic processing task was also measured only at the age of 13 years. This fact limits our ability to determine cause and effect. However, this study provides preliminary data for the cooccurrence of L1 and L2 word reading difficulties. The findings indicate that students who were diagnosed a few years earlier as having dyslexia demonstrated word reading difficulties rather consistently, as well as subsequent difficulties with L2 English orthographic processing. This points to the stability of the dyslexic profile of Chinese children who are learning to read concurrently in Chinese and English. This could guide the development of future studies on potential L2 reading and orthographic processing difficulties in children with L1 dyslexia. It might be of both theoretical and practical importance to expand the current focus on L2 word reading difficulty to L2 reading difficulty at the sentence and discourse levels, which were not included in this study. A third limitation concerns the design of the stimuli in the English orthographic processing task. In this study, the English orthographic task included real word, look-like words, and nonwords. Although we controlled for some orthographic constraints such as consonant doublets at the beginning of words and the fact that vowel doublets were not allowed across all three conditions, we did not match the stimuli according to statistically derived orthographic/phonotactic probabilities or neighborhood density. It would be challenging to match stimuli according to these properties, given that the English curriculum for Hong Kong is different from what it is for L1 speakers, and we have no clear measures of these constructs for Hong Kong Chinese children. Nevertheless, these are important issues to consider in future research. Finally, in this study we measured orthographic processing only in L2 English. We had a clear focus on English here, but it may be of interest to examine how L1 Chinese orthographic processing correlates with L2 English orthographic processing and the extent to which both contribute to L1 and L2 word reading in dyslexic readers within and cross-linguistically. Despite these limitations, this study provides empirical evidence for the hypothesis that children with dyslexia in Chinese manifest L2 word reading difficulties in English as well. It is possible that Hong Kong

180 TOPICS IN LANGUAGE DISORDERS/APRIL JUNE 2017 Chinese children diagnosed with dyslexia have unusual difficulties in orthographic analysis. It looks as though these children have particularly strong difficulties in tasks that require with orthographic pattern recognition of look-like words. This difficulty with the English orthographic awareness measure is particularly interesting, given our finding that Chinese children with dyslexia only had significantly lower English real-word reading at ages of 11 and 12 years. Indeed, at younger ages, the difference in L2 English word reading between dyslexic children and normal readers is not apparent; this might be because they are all beginning readers in L2. This hypothesis is worth further exploring in future work. On the whole, this study shows that it is not unusual for children who are dyslexic in their L1 to have difficulties in L2 word reading at ages of 10 and 11 years. Not all children with dyslexia in Chinese showed a pattern of problems with English as well, but many of them did. REFERENCES Bernhardt, E. B., & Kamil, M. L. (1995). Interpreting relationships between L1 and L2 reading: Consolidating the linguistic threshold and the linguistic interdependence hypotheses. Applied Linguistics, 16(1), 15 34. Burt, J. S. (2006). What is orthographic processing skill and how does it relate to word identification in reading? [Article]. Journal of Research in Reading, 29(4), 400 417. doi:10.1111/j.1467-9817.2006.00315.x Cassar, M., & Treiman, R. (1997). The beginnings of orthographic knowledge: Children s knowledge of double letters in words. Journal of Educational Psychology, 89(4), 631. Chung, K. K. H., & Ho, C. S. H. (2009). Second language learning difficulties in Chinese children with dyslexia: What are the reading-related cognitive skills that contribute to English and Chinese word reading? Journal of Learning Disabilities, 43(3), 195 211. Coltheart, M., Rastle, K., Perry, C., Langdon, R., & Ziegler, J. (2001). DRC: a dual route cascaded model of visual word recognition and reading aloud. Psychological review, 108(1), 204. Corcos, E., & Willows, D. M. (1989). A developmental study of the processing of orthographic information in children with varying reading ability. Paper presented at the Annual Meeting of the National Reading Conference, Austin, TX. Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual children. Review of Educational Research, 49(2), 222 251. Cummins, J. (1984). Bilingualism and special education: Issues in assessment and pedagogy (Vol. 6). Clevedon, England: Multilingual Matters. Education Department. (1996). Guide to the pre-primary curriculum. Hong Kong: Government Printer. Geva, E., & Siegel, L. S. (2000). Orthographic and cognitive factors in the concurrent development of basic reading skills in two languages. Reading and Writing, 12(1), 1 30. Hanley, J. R., Hastie, K., & Kay, J. (1992). Developmental surface dyslexia and dysgraphia: An orthographic processing impairment. The Quarterly Journal of Experimental Psychology Section A, 44(2), 285 319. doi:10.1080/02724989243000046 Ho, C. S.-H., Chan, D. W., Tsang, S., & Lee, S. (2000). The Hong Kong Test of Specific Learning Difficulties in Reading and Writing (HKT-SpLD) manual. Hong Kong: Hong Kong Specific Learning Difficulties Research Team. Ho, C. S.-H., Chan, D., Chung, K., Tsang, S.-M., Lee, S.-H., & Cheng, R. W.-Y. (2007). The Hong Kong Test of Specific Learning Difficulties in Reading and Writing for Primary School Students-Second Edition [HKT-P(II)]. Hong Kong: Hong Kong Specific Learning Difficulties Research Team. Ho, C.-H., & Fong, K.-M. (2005). Do Chinese dyslexic children have difficulties learning English as a second language? Journal of Psycholinguistic Research, 34(6), 603 618. doi:10.1007/s10936-005-9166-1 Koda, K., 2005. Insights into second language reading: A cross-linguistic approach. Cambridge University Press, Cambridge. Lyon, G. R., Shaywitz, S., & Shaywitz, B. (2003). A definition of dyslexia. Annals of Dyslexia, 53(1), 1 14. doi:10.1007/s11881-003-0001-9 McBride-Chang, C., Shu, H., Chan, W., Wong, T., Wong, A. M. Y., Zhang, Y., et al. (2012). Poor readers of Chinese and English: Overlap, stability, and longitudinal correlates. Scientific Studies of Reading, 17(1), 57 70. doi:10.1080/10888438.2012.689787 McBride-Chang, C., Tong, X., Shu, H., Wong, A. M. Y., Leung, K. W., & Tardif, T. (2008). Syllable, phoneme, and tone: Psycholinguistic units in early Chinese and English word recognition. Scientific Studies of Reading, 12(2), 171 194. Perfetti, C. A., & Hart, L. (2002). The lexical quality hypothesis. Precursors of Functional Literacy, 11, 67 86. Perfetti, C. A., Liu, Y., Fiez, J., Nelson, J., Bolger, D. J., & Tan, L. H. (2007). Reading in two writing systems: Accommodation and assimilation of the brain s

English Word Reading Difficulties and Orthographic Processing Weaknesses 181 reading network. Bilingualism: Language and Cognition, 10(02), 131 146. Raven, J. C., Court, J. H., & Raven, J. (1995). Raven s coloured progressive matrices. Oxford, United Kingdom: Oxford Psychologists Press. Seidenberg, M. S. (2011). Reading in different writing systems: One architecture, multiple solutions. P. McCardle, B. Miller, J. Lee & O. Tzeng. Dyslexia across languages: orthography and the brain-gene-behavior Link (pp. 151-174). Baltimore: Brookes. Snowling, M. (1998). Dyslexia as a phonological deficit: Evidence and implications. Child Psychology and Psychiatry Review, 3(1), 4 11. doi:10.1111/1475-3588.00201 Stanovich, K. E., & West, R. F. (1989). Exposure to print and orthographic processing. Reading Research Quarterly, 24(4), 402 433. Tong, X., & McBride-Chang, C. (2010). Chinese English biscriptal reading: Cognitive component skills across orthographies. Reading and Writing, 23(3/4), 293 310. Tong, X., Tong, X., & McBride-Chang, C. (2015). A tale of two writing systems: Double dissociation and metalinguistic transfer between Chinese and English word reading among Hong Kong children. Journal of Learning Disabilities, 48(2), 130 145. doi:10.1177/0022219413492854 Treiman, R. (1993). Beginning to spell: A study of first grade children. New York: Oxford University Press. Van Gelderen, A., Schoonen, R., De Glopper, K., Hulstijn, J., Simis, A., Snellings, P., et al. (2004). Linguistic knowledge, processing speed, and metacognitive knowledge in first- and second-language reading comprehension: A componential analysis. Journal of educational psychology, 96(1), 19. Wade-Woolley, L., & Geva, E. (2000). Processing novel phonemic contrasts in the acquisition of L2 word reading. Scientific Studies of Reading, 4(4), 295 311. Zhou, Y., McBride-Chang, C., Law, A. B.-Y., Li, T., Cheung, A. C.-Y., Wong, A. M. Y., & Shu, H. (2014). Development of reading-related skills in Chinese and English among Hong Kong Chinese children with and without dyslexia. Journal of Experimental Child Psychology, 122, 75 91. doi:http://dx.doi.org/10.1016/j. jecp.2013.12.003