三加一的幸福方程式
關於部落格
赤子之心 永相隨
  • 3428

    累積人氣

  • 0

    今日人氣

    0

    訂閱人氣

詞素覺識

This 1-year longitudinal study examined the extent to which morphological awareness,orthographic knowledge, and phonological awareness, along with speeded naming, uniquely explained word recognition, dictation (i.e., spelling), and reading comprehension among 171 young Hong Kong Chinese children. With age and vocabulary knowledge statistically controlled, both morphological awareness and orthographic knowledge were uniquely associated with all three concurrently measured literacy skills, as well as longitudinal measures of specific literacy skills. Naming speed was also uniquely associated with concurrent word reading, as well as all three literacy skills longitudinally, even with their autoregressive effects controlled. Analyses of children’s spelling mistakes indicated that 97% and 95% of all errors were either morpholexically or orthographically based at times 1 and 2, respectively. Morphologically based spelling errors were also uniquely associated with all three literacy skills across time. Findings underscore the importance of morphological awareness and orthographic knowledge for Chinese literacy acquisition. Correspondence should be sent to Catherine McBride-Chang,Department of Psychology, Chinese University of Hong Kong, Shatin, New Territories, Hong Kong. E-mail: cmcbride@psy.cuhk.edu.hk MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 427 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 A growing number of studies on Chinese literacy skills have either implicitly or explicitly contrasted Chinese and alphabetic language reading and writing development (e.g., Ho, Yau, & Au, 2003; Leong, Tse, Loh, & Hau, 2008; McBride-Chang, Tong, et al., 2008; Shu, McBride-Chang, Wu, & Liu, 2006; Wang, Cheng, & Chen, 2006). Although not discounting the importance of phonological awareness for young Chinese children’s literacy skills development (e.g., Ho & Bryant, 1997b; McBride-Chang & Ho, 2000), they nevertheless highlight the importance of morphological awareness (McBride-Chang, Shu, Zhou, Wat, & Wagner, 2003; Shu et al., 2006) and, to a lesser extent, orthographic skills (e.g., Ho & Bryant, 1997a; Ho et al., 2003; Siok & Fletcher, 2001), for Chinese literacy acquisition, even among relatively young children. The present study expanded this work by targeting four reading-related skills that have been of primary, though by no means equal, focus in much of the cross-cultural work on reading acquisition, namely, speeded naming, visual-orthographic skills, phonological awareness, and morphological awareness. This study was a 1-year longitudinal study targeting three core literacy skills in young Chinese children, that is, word recognition, word dictation (spelling), and reading comprehension. Our overarching hypothesis was that morphological and orthographic skills would emerge as important both concurrently and longitudinally for explaining development across these literacy skills, whereas the role of phonological awareness for learning to read and write in Chinese would be relatively small. We expected that the greater focus on morphological awareness and orthographic skill relative to phonological awareness would also be reflected in the errors Chinese children made in their spelling, or dictation,1 of Chinese words across the two testing times. Given the importance of speeded naming for Chinese reading development (e.g., McBride-Chang, Tong, et al., 2008) and impairment (e.g., Ho & Lai, 1999; Shu et al., 2006), this measure was also included in the present study, though its theoretical relevance for Chinese literacy skills was not a central focus. We briefly review the literature on each of our four targeted reading-related skills and then highlight why morphological awareness and orthographic knowledge may be particularly interesting and important to consider for Chinese developing readers. Studies of Chinese children’s reading acquisition and development have demonstrated that speeded naming (e.g., Ho & Lai, 1999), visual-orthographic skills (e.g., Ho et al., 2003; Huang & Hanley, 1997; Siok & Fletcher, 2001), phonological awareness (e.g., Ho & Bryant, 1997b; McBride-Chang & Ho, 2000), and morphological awareness (W. L. Li, Anderson, Nagy, & Zhang, 2002; McBride- Chang et al., 2003; Shu et al., 2006; Wang et al., 2006) are important for Chinese 1Because Chinese words are written using characters, rather than letters, there is no English equivalent word that precisely defines the process by which children produce Chinese words they are asked to write. Therefore, the terms dictation and spelling are both used interchangabley throughout this paper to represent this idea. 428 TONG ET AL. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 word reading development and impairment. Similar skills are associated with word recognition across alphabets (e.g., Adams, 1990; Deacon, Wade-Woolley, & Kirby, 2007; McBride-Chang, 2004; Snowling, 2000). In part because of the relatively arbitrary nature of sound–symbol correspondences of Chinese, speeded naming is one of the strongest correlates of Chinese word recognition and among the two most dominant types of cognitive deficits found in Hong Kong Chinese children with dyslexia so far (e.g., Ho, Chan, Lee, Tsang, & Luan, 2004; Ho, Chan, Tsang, & Lee, 2002). Visual-orthographic skills also tend to be strongly related to reading impairment in young Chinese children (e.g., Ho et al., 2004; Ho et al., 2002), though relatively few studies have included such measures in studies of typically developing children, as we did in the present study. Here, visual-orthographic skill was defined as children’s awareness of conventional rules in structuring Chinese characters and to identify or distinguish real Chinese characters from a set of pseudocharacters, noncharacters, and visual symbols. Given the universality of reading as a pairing of phonological information with visual information, the importance of phonological awareness for early reading acquisition in Chinese is generally acknowledged as well, though the relatively simple phonological structure of Chinese language may reduce the importance of phonological awareness for reading Chinese in older children (e.g., McBride-Chang, Bialystok, Chong, & Li, 2004). Finally, morphological awareness as measured via lexical compounding and homophone awareness in relation to Chinese words appears to be important for early word recognition in Chinese children because of the nature of the Chinese script. Note that the way in which morphological awareness has been measured in Chinese children thus far has differed substantially from the focus on inflectional or derivational morphology typically examined in studies of children learning to read alphabetic scripts (e.g., Deacon et al., 2007; Nagy, Berninger, & Abbott, 2006). There are at least two plausible reasons for positing an underlying connection of Chinese morphological awareness to Chinese word reading, spelling, and reading comprehension. First, as outlined in the metalinguistic processing model of Bialystok and Ryan (1985), analyzed knowledge and cognitive control of linguistic processing are the key skill components underlying both metalinguistic development and, ultimately, literacy development. Literacy development itself further propels children’s growth in the ability to analyze and manipulate language (Chomsky, 1979; Leong, Hau, Cheng, & Tan, 2005; Nagy, Berninger, & Abbott, 2003). Hence, the ability to access or manipulate morphemic structures flexibly and accurately, a basic definition of morphological awareness (Carlisle, 1995), in some ways parallels the broader concept of analyzed knowledge (Bialystok & Ryan, 1985) and its links to word reading, spelling, and reading comprehension. MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 429 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 Second, Chinese can be operationally defined as a meaning-centralized writing system. For example, the Chinese character, the basic written unit, is a convergence of sound, meaning, and form. Because a fundamental processing task underlying reading, spelling, and reading comprehension in Chinese is mental representation or integration of meaning, literacy skills draw heavily upon Chinese children’s knowledge of morphemes or morphemic structure. Thus, for both theoretical and practical reasons, morphological awareness is hypothesized to be important across Chinese literacy skills from very early on in development. Given that the importance of lexical compounding and homophone production, two core components of morphological awareness in Chinese character recognition, we sought to test these ideas by examining the contributions of morphological awareness and other metalinguistic skills to Chinese children’s spelling skills. We tested these ideas by examining both the associations across metalinguistic abilities to children’s total spelling score and children’s spelling errors within a spelling test. Children’s spelling errors have been viewed as a window into children’s orthographic knowledge development and strategy shift (Shen & Bear, 2000). Such spelling errors are not randomly distributed but follow a pattern (Read, 1986). For example, Shen and Bear showed that Chinese children’s invented spelling errors were generally composed of phonological-, graphemic-, and morphological-based errors and there existed a clear developmental shift of these three types of errors, progressing from primarily phonological to orthographic to morphological, as children’s grade level increased from 1 to 6 (Shen & Bear, 2000). Ho et al. (2003) also proposed an important developmental model of orthographic knowledge development in relation to spelling based on children in kindergarten through second grade. However, both studies focused primarily on analyzing spelling at the character level only. In the present study, we focused on young Hong Kong Chinese children’s spelling by instead analyzing errors at the word level. A word in Chinese typically consists of two or more characters (e.g., Shu, Chen, Anderson, Wu, & Xuan, 2003). Given the many homophones and extensive lexical compounding of Chinese, the spellings of the two characters in a given Chinese word are often mutually cued in the specific context. To test the importance of word-level spelling, we examined errors across all two-morpheme words included on a spelling test to see how many occurred at the level of the character and how many at the level of the word. We also tested the importance of each metalinguistic skill, including morphological awareness, for explaining variance in overall spelling performance (as well as word recognition performance), both concurrently and longitudinally. We further examined morphological awareness and other metalinguistic skills in relation to reading comprehension, the ultimate indicator of literacy skills. At least three studies of Chinese children’s reading comprehension (Leong, Hau, Tse, & Loh, 2007; Leong et al., 2008; Shu et al., 2006) have tested components of reading comprehension in Chinese children. One (Leong et al., 2007) demonstrated that ver 430 TONG ET AL. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 bal memory and pseudoword reading were particularly salient in distinguishing good from poor comprehenders in Hong Kong; another (Leong et al., 2008) underscored the particular importance of verbal memory for modeling reading comprehension in a large sample of third-through fifth-grade Hong Kong Chinese children. The third (Shu et al., 2006) showed that morphological production, speeded naming, and vocabulary skills most clearly distinguished readers with dyslexia from those without reading impairments and explained unique variance in different reading skills, including reading comprehension, among Mainland (Beijing) Chinese children in Grades 5 and 6. As compared to previous work in this area, the present study focused on diverse linguistic/cognitive correlates of reading comprehension in much younger children developing over one year. In this work, we tested a typically developing sample of Hong Kong Chinese children in kindergarten and first grade. Unlike the previous studies that made use of production (e.g., filling in the blank, answering a question) in writing to tap reading comprehension, we tested reading comprehension individually in our children using a sentence selection strategy to maximize the applicability of our measures for young children. Although the children in the present study were clearly young to be tested on reading comprehension by Western standards, they had all been formally trained in reading skills for 2.5 to 3.5 years by the time they were tested here. Their primary focus in class had become reading comprehension at the sentence and paragraph level. To test the importance of a variety of metalinguistic skills for literacy development in a comprehensive way, therefore, sentence-level reading comprehension performance was measured in the present study. To summarize, in the present study, we tested the concurrent and longitudinal associations of four metalinguistic skills, speeded naming, visual-orthographic ability, phonological awareness, and morphological awareness, in relation to word recognition, spelling, and reading comprehension among young Hong Kong Chinese children. We were particularly interested in the importance of morphological awareness relative to other metalinguistic abilities in relation to all three literacy skills across time. Therefore, we analyzed the distributions of children’s spelling errors to identify the extent to which children’s early spelling strategies tended to be morphologically, as well as phonologically and orthographically, based over the 2 years tested. Finally, we examined how each of these three strategies was concurrently and longitudinally associated with word reading, word spelling, and reading comprehension performance. METHOD Participants Participants were 196 (109 girls, 87 boys) Hong Kong Chinese children who were attending the third year of kindergarten in Hong Kong at the beginning of the MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 431 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 study. Twenty-five of these (14 girls, 11 boys) had dropped out by Time 2. Thus, the final analyses reported here consisted of 171 (98 girls, 73 boys) children who were tested twice over a 1-year period, with a mean age at the first testing time of 6.13 years (range = 5.50 to 6.67, SD = .30 years). The participants were all from Cantonese-speaking homes and attending Cantonese-medium schools. Family incomes reported by parents of these children were in the range of median household income levels according to the 2006 census of the Hong Kong population (Hong Kong Census and Statistics Department, 2006). In addition, all the children were reported by parents to be typically developing, without any identified developmental delays. Children were selected from across Hong Kong, so they attended many different schools. Typically, Hong Kong Chinese children begin reading instruction around the ages of 3 to 4 years old and are taught to write around the ages of 3 to 5 years old. According to parental reports, all children had begun literacy instruction within this time period in the present study. Measures Chinese character recognition. A Chinese character test adopted from the Hong Kong Test of Specific Learning Difficulties in Reading and Writing (Ho, Chan, Tsang, & Lee, 2000) consisting of 150 two-character words, ranked according to increasing difficulty, was administered at both Times 1 and 2. Children were required to read aloud the words from left to right and top to bottom. The testing was stopped when children failed to read fifteen consecutive items. The maximum score on this task was 150. Chinese word dictation. The Chinese word dictation task assessed children’s word spelling skills at both Time 1 and 2. It was composed of 20 and 25 two-character words adopted from the Hong Kong Test of Specific Learning Difficulties in Reading and Writing (Ho et al., 2000) for Time 1 and Time 2, respectively. All the words to be spelled on this task were representative of beginning words taught in Hong Kong. All the items were in order of increasing difficulty. In the testing, the experimenter read aloud the words and children were asked to write each in the blank space. The testing was stopped when children missed five consecutive words in a row. One mark was given for each correctly written two-character word, and the maximum score of this task was 20 at Time 1 and 25 at Time 2. Chinese reading comprehension. Chinese reading comprehension was assessed twice in this study. Given these children’s young ages and reading experiences, this newly designed Chinese reading comprehension test was created to tap children’s ability in understanding the content and the structure of a very short passage. It consisted of 36 and 30 items, respectively, at Times 1 and 2 (18 items overlapped across time), and the difficulty levels of these items spanned from typical 432 TONG ET AL. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 kindergarten to typical second grader skills according to pilot testing data. However, after analyzing individual items at Time 1, some of the items were changed to maximize breadth of testing. All items for testing Times 1 and 2 were separated into three blocks in terms of difficulty levels, with 10 to 12 items per block. Each item consisted of a short print narrative that described a specific number of objects, persons, or animals, or a sequential event, action, or story, and five corresponding line-drawing pictures. Of these, each narrative generally consisted of between one and two simple sentences, and each simple sentence consisted of single and compound words, ranging from 5 to 20. The texts of these items involved processes of coordination, subordination, modification, and cause–effect. All of these narratives represented common events or activities children encountered in daily life. Neither individual word knowledge nor basic background (world) knowledge could by itself facilitate children’s understanding of the whole passage. The directionality, causation, or relationship underlying the text could be inferred only from the integration of key words in the text, or building a mental model of the whole passage. Of the five pictures, one was the target picture, representing the exact contents or scene described in the printed narrative. The other four pictures were distracters that included some, but not all, information of the printed text. Participants were visually presented with a printed narrative such as “A cat stands inside the shoe,” along with five pictures including (a) a cat pushing a shoe, (b) a mouse looking at the shoe, (c) the cat inside the shoe (correct answer), (d) a mouse inside the shoe, and (e) a cat looking at the shoe. Children were asked to select, from among these five pictures, the one best representing the content described by the text and to circle the number corresponding to the picture. All children began at the beginning of the task, and testing was stopped when children failed in five consecutive items within a given block. The maximum expected score for the kindergarteners was 12 based on pilot testing for Time 1. However, the maximum total score of this task was 36 at Time 1 because of the plan to extend this task to older children. Item analyses of Time 1 led us to some changes in items at Time 2, though the basic structure of the test remained the same across testing times. Some sample items used in this task are shown in the appendix. Vocabulary definitions. A Cantonese vocabulary definitions task was used to assess children’s general vocabulary knowledge at Time 1 only. It consisted of 53 words arranged in order of ascending difficulty level, and it has been successfully used in previous studies as an indicator of vocabulary knowledge in Hong Kong Chinese children (McBride-Chang, Tardif, et al., 2008). All of these items were selected from Hong Kong Chinese children’s reading books (Zhuang, 2000), and they were pilot-tested on children of the same age. Participants were orally presented with a word representing an object or concept, and the children were asked to orally explain or define this word. Children’s answers were scored with reference to crite MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 433 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 ria similar to those used in the Stanford-Binet Intelligence Scale vocabulary subtest (Thorndike, Hagen, & Sattler, 1986). A scoring scheme and sample answers for each item was developed on the basis of the description provided by a Chinese dictionary (Lau, 1999). Sample answers for scores of 0, 1, and 2 were also included in the marking scheme. Specifically, 2 points were given for a synonym or a clear description of the meaning or the function of the target word. One point was allotted for an ambiguous or incomplete answer, or an example of the target word. A score of zero was given for repetition of the target word, or the irrelevant explanation of the target word. Testing was stopped when the children scored zero on five consecutive items. The possible maximum score of this task was 106. Speeded naming (of numbers). The speeded naming of numbers was measured using a previously developed test (e.g., McBride-Chang, Tong, et al., 2008) at Time 1 only. It is made up of five rows of 5 single-digit numbers, with a total of 25 numbers, randomly arranged on a single sheet. Each row consisted of the same 5 single-digit numbers presented in a different order. Children were first asked to name each number slowly in order to ensure their familiarity with all of these numbers, and then they were required to name all the digits as rapidly as possible in Cantonese. All of these numbers were named twice, and the average of these two scores was taken to represent the total score on this task. Phonological awareness. We created a phonological awareness task in the spirit of the Comprehensive Test of Phonological Processing (Wagner, Torgesen, & Rashotte, 1999) in which children were tested on three-syllable words or phrases, both real and nonsense, conforming to the constraints of Cantonese. In this section, children were asked first to delete the first, then the middle, and then the final syllable (with 10 items for both the first and middle positions, and 9 items for the final position). Once these were completed, children were asked to delete an initial sound (phoneme) from a single syllable (e.g., si1 without the initial phoneme would be i1), beginning with real words and moving onto nonsense words (11 each). Following extensive testing, we set basal rules for this task such that if children failed in one item or less for a given block, they advanced to further levels and were not tested on any previous items before these. The ceiling set rule was that testing stopped when children got 6 consecutive items wrong for anywhere in the syllable deletion section or 4 consecutive items incorrect for the phoneme deletion section. This test was used successfully in a previous study of Hong Kong Chinese children who were disabled and nondisabled readers (Chung et al., 2008), and it was administered at Time 1 only for the present study. Testing began at the beginning of this task for all children. The maximum score of this task was 51. 434 TONG ET AL. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 Morphological awareness. Morphological awareness was tested using morphological construction and homophone knowledge items ranked according to increasing difficulty level with reference to pilot testing data from a large sample of children spanning from kindergarten to fifth grade, similar to the process used for the phonological awareness measure just described. This test was also used in previous research distinguishing Chinese children with and without dyslexia (Chung et al., 2008), and it was administered at Time 1 only. The first part of the task required children to combine familiar morphemes to make up 27 new concepts or novel words. Participants were orally presented a three-sentence scenario that described a concept or object orally to children for each item, and children were then required to come up with a compound word to represent it based on the scenario. To aid children’s understanding of the task, two trial items were illustrated using pictures. The test items are relatively abstract without any accompanying picture. For example, one testing item was “The scene we saw in the night from the top of the mountain is called a night scene. What would we call the scene that is seen in the morning from the top of the mountain?” The correct answer would be “morning scene.” Children began this test from the beginning, and if they missed four or more consecutive items on this section of the task, testing stopped. For those who advanced all the way through these items, a second part of the test, focused on homophone production, was administered. In this section, for each item, one target monosyllable or morpheme was orally presented by the experimenter in a word context. Children were first required to use the target morpheme to construct a different word making use of the same morpheme within 20 sec. Afterward, children were asked to name another word that made use of a homophone morpheme, that is, that used the homophone in a different word with a different meaning. This task has been used previously in older children (e.g., Shu et al., 2006). An example in English would be to come up with beehive and become for the target bee in honeybee. Beehive makes use of the same morpheme (bee) as in the target word, whereas become makes use of the homophone of the target morpheme (be). A real example from Cantonese was that, given the target item /syu1/ (book) in the word bookbag, children were asked to come up with some new compound word such as textbook, bookshop, bookshelf, bookwriting and bookmark. Next, children were asked to give as many words that included the homophone of /syu1/ (book) as possible, but the homophone was required to have a distinct meaning and form from the initial , such as /syu1 fuk6/ (comfortable), /syu1 tsing4/ (expression), or /syu1 ying4/ (win–lose). This part of the task requires a relatively high level of processing. Because this was a production exercise administered within a limited time, no maximum score was specified for this section. All items were administered for this second part of the test once children attained this level. MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 435 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 Orthographic knowledge. A Chinese character decision task was created to assess children’s awareness of the internal structure of the Chinese character, and their ability to identify or distinguish real Chinese characters from a set of pseudocharacters, noncharacters, and visual symbols at Time 1 only. It was composed of 14 sets of 5 items each, for a total of 70 items arranged in a paper booklet. Of these, there were 30 real characters, 9 pseudocharacters, 21 noncharacters, and 10 visual symbols. Moreover, the high-, middle-, and low-frequency real Chinese characters were evenly distributed across items. The pseudocharacters were created by forming a novel combination of two Chinese radicals or components conforming to Chinese character constraints. Noncharacters were designed by subtracting or adding strokes or switching the position of the components of real Chinese characters, violating the conventional rules of Chinese characters. Visual symbols consisted of a variety of graphic or simple line drawings. Each set was composed of pseudocharacters, noncharacters, and visual symbols, and they were randomly arranged on a single sheet. In the testing, each set of 5 items was presented to children, and the children were required to indicate whether each item was a real character or noncharacter. One mark was scored when children correctly identified a real character or a noncharacter, and the maximum possible score for this task was 70. Procedure Consent was obtained from parents and children for this study and a convenient testing time at home was arranged for all the children. Participants were tested on tasks of Chinese character recognition, Chinese word dictation, Chinese reading comprehension, vocabulary definition, speeded naming, phonological awareness, morphological awareness, and orthographic knowledge at the beginning of this study. All of these tests were individually administered to children in two 50-min sessions by five trained undergraduate psychology majors in a quiet room at home. A 15-min break was given between the two sessions, and shorter breaks were given as needed (i.e., when children requested breaks or looked as though they needed one). The same children were tested again on tasks of Chinese character recognition, Chinese word dictation, and Chinese reading comprehension 1 year later. The orders of the tasks were counterbalanced across participants. RESULTS Means, standard deviations and reliabilities of all the measures administered at Times 1 and 2 are shown in Table 1. The reliability of speeded naming shown in the table is test–retest reliability, whereas the reliabilities for all other tasks are internal consistency reliabilities. As noted in the table, the reliabilities of all the read Downloaded By: [University of Alberta] At: 08:44 19 December 2009 436 TONG ET AL. TABLE 1 Means, Standard Deviations, and Reliabilities for All the Measures Across Time Measures (Maximum Score Possible) M SD Reliability Chinese character recognition, T2 (150) 81.56 28.40 .98 Chinese word dictation, T2 (25) 10.22 5.93 .89 Chinese reading comprehension, T2 (30) 17.20 5.23 .79 Chinese character recognition, T1 (150) 33.45 28.84 .99 Chinese word dictation, T1 (20) 4.94 3.27 .83 Chinese reading comprehension, T1 (36) 8.09 6.70 .94 Vocabulary definitions,T1 (106) 21.29 8.64 .88 Speeded naming,T1 (—) 14.41 4.40 .92 Phonological awareness, T1 (51) 26.57 8.05 .94 Morphological awareness, T1 (—) 17.00 9.45 .97 Orthographic knowledge, T1 (70) 45.29 6.49 .98 Note. N = 171. All reliabilities listed are internal consistency reliabilities except for the speeded naming task, which was measured as test–retest reliability. A dash indicates that there were no maximum possible scores for the speeded naming and production tasks. T1 = Time 1; T2 = Time 2. ing-related tasks included in this study were above .75. In addition, the mean of the reading comprehension at Time 1 was 8.09 out of 36. Although this is relatively low in terms of total score, given that the 36 items in this task were designed for children from kindergarten to second grade, so that, of these, 12 items were supposed to match the average reading proficiency of the test for kindergartners only, this mean score is reasonable. Moreover, a histogram revealed that the children’s performances were relatively normally distributed for this task. Correlations among phonological awareness, morphological awareness, orthographic knowledge, and reading outcome measures across time are shown in Table 2. As predicted, correlations involving Chinese character recognition, word dictation, and reading comprehension were strong, ranging from .49 to .62 at Time 1. There was a similar association pattern for these three literacy skills measured at Time 2, with correlations ranging from .41 to .61. Moreover, cross-lag (Times 1 and 2) correlations for Chinese character recognition, word dictation, and reading comprehension across time were modest, ranging from .29 to .79. All correlations of phonological awareness and morphological awareness with the three literacy skills were moderately significant across time. Correlations of orthographic knowledge, with Chinese character recognition, word dictation, and reading comprehension were also moderately significant across time, ranging from .20 to .37. Speeded naming was also significantly negatively correlated with Chinese literacy skills and other measures across time, demonstrating that faster children tended to perform better across literacy tasks across time. The vocabulary definitions task was also significantly associated with all other variables across time, except for Chinese dictation at Time 2. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 TABLE 2 Correlations Among Phonological Awareness, Morphological Awareness, Orthographic Knowledge, Vocabulary Definitions, Speeded Naming, and Literacy Skill Measures Variables 1 2 3 4 5 6 7 8910 11 Chinese character recognition, T2 1. — Chinese word dictation, T2 2. .56*** — Chinese reading comprehension, T2 3. .61*** .41*** — Chinese character recognition, T1 4. .79*** .40*** .51*** — Chinese word dictation, T1 5. .56*** .56*** .41*** .60*** — Chinese reading comprehension, T1 6. .53*** .29*** .41*** .62*** .49*** — Vocabulary definitions, T1 7. .39*** .15. .36*** .40*** .26** .29*** — Speeded naming, T1 8. –.49*** –.33*** –.43*** –.41*** –.34*** –.29*** –.29*** — Phonological awareness, T1 9. .28*** .32*** .34*** .26** .33*** .22** .29*** –.38*** — Morphological awareness, T1 10. .44*** .26** .44*** .52*** .41*** .42*** .48*** –.36*** .54*** — Orthographic knowledge, T1 11. .37*** .20** .25** .42*** .36*** .35*** .18* –.28*** .12 .25** — Note. T1=Time 1;T2 =Time 2. . p < .10. *p < .05. **p < .01. ***p < .001. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 438 TONG ET AL. Explaining Concurrent and Longitudinal Literacy Skills To examine the unique contributions of phonological awareness, morphological awareness, and orthographic knowledge at Time 1 to concurrent and subsequent measures of character recognition, word dictation, and reading comprehension, a set of multiple regression analyses explaining Chinese character recognition, word dictation, and reading comprehension was conducted separately for Time 1 and 2. Each set of analyses contained the vocabulary definitions test and age in the first step in each equation as controls. Speeded naming was entered at Step 2 given that it has been shown to be important for Chinese children’s word reading ability (e.g., McBride-Chang, Tong, et al., 2008). In Step 3, phonological awarenes, orthographic knowledge, and morphological awareness were included to investigate the unique and shared contributions of each for different literacy outcome measures across time. The concurrent relations of predictor and literacy outcome variables (Time 1 measures to Time 1 word reading, dictation, and reading comprehension) and longitudinal ones (Time 1 measures to Time 2 word reading, dictation, and reading comprehension) are displayed in Table 3 including final standardized betas (β), semipartial correlations (sr2i) and total R2. A primary focus was to examine whether morphological awareness would make a unique contribution to each literacy skill measured concurrently and subsequently in the present study. Morphological awareness consistently explained concurrent Chinese word reading (sr2i = .07), word dictation (sr2i = .02) , and reading comprehension (sr2i = .05) and its longitudinal relations with Chinese word reading (sr2i = .02), and reading comprehension (sr2i = .02), but not with word dictation, were also significant, even after controlling for age, vocabulary, and other reading-related measures, as reported in Table 3. As is clear from Table 3, orthographic knowledge also appeared to be a salient factor relating to early Chinese literacy skills in the present study. Orthographic knowledge significantly explained concurrent Chinese word reading (sr2i = .06), word dictation (sr2i = .04), and reading comprehension (sr2i = .04) with age, vocabulary, and speeded naming statistically controlled. However, only the longitudinal association of orthographic knowledge with word reading (sr2i = .04) was significant. In contrast, phonological awareness was uniquely associated with subsequent word dictation only (sr2i = .03), and it failed to explain unique variance in the other literacy measures across time. In addition, although speeded naming made a unique contribution to concurrent Chinese word reading only, its effects on subsequent word reading, word dictation, and reading comprehension were relatively strong even controlling for age and vocabulary. In a more stringent test of the associations of morphological awareness, orthographic knowledge, and phonological awareness to the longitudinal measures of word reading and word dictation, as well as reading comprehension, we also controlled for Time 1 word reading, Time 1 word dictation, or Time 1 reading compre Downloaded By: [University of Alberta] At: 08:44 19 December 2009 TABLE 3 Hierarchical Regression Predicting Chinese Character Recognition, Chinese Word Dictation and Reading Comprehension Across Time From Age, Vocabulary Definitions, Speeded Naming, Phonological Awareness, Morphological Awareness, and Orthographic Knowledge Measured at Time 1 Chinese Character Recognition Chinese Word Dictation Chinese Reading ComprehensionConcurrent Subsequent Concurrent Subsequent Concurrent Subsequent Step Variable sr2 sr2 sr2 sr2 sr2 sr2 1. Vocabulary definitions, T1 .136 .014. .174 .022* –.009 –.000 –.019 .000 .038 .001 .137 .014. Age, T1 .031 .001 –.016 .000 .197 .034** –.013 .000 .213 .040** .028 .001 2. Speeded naming, T1 –.193 .029** –.318 .077*** –.126 .012. –.211 .034 –.090 .006 –.258 .051** 3. Phonological awareness, T1 –.076 .004 –.023 .000 .134 .012. .197 .026* –.030 .001 .083 .005 Orthographic knowledge, T1 .255 .057**.206 .037* .214 .040** .108 .010 .200 .035** .084 .006 * Morphological awareness, T1 .353 .070**.204 .023* .196 .022* .060 .002 .288 .047** .204 .023* * R 2(df) .41 (3, 164) .38 (3, 164) .30 (3, 164) .17 (3, 164) .29 (3, 164) .30 (3, 164) Note. T1 = Time 1. N= 171. . p < .10. *p < .05. **p < .01. ***p < .001. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 440 TONG ET AL. hension as corresponding autoregressors of subsequent outcome measures, respectively. In these analyses, speeded naming was the only significant predictor of Chinese word reading (sr2i = .02), word dictation (sr2i = .02), and reading comprehension (sr2i = .04) after statistically removing the effects of age, vocabulary, and the respective autoregressor. Across these longitudinal equations, 67% of the variance in Chinese character recognition, 36% of the variance in Chinese word dictation, and 33% of the variance in reading comprehension was explained by these variables combined. Analyses of Spelling Errors In addition to these regression equations, we analyzed children’s spelling errors individually for patterns in two-morpheme words. Following Shen and Bear’s (2000) analysis of Chinese invented spelling errors, phonological-based errors, orthographic- based errors, and morpho-lexical errors were generally categorized in terms of the deviation of children’s spelling from the character in the target word. Of all 171 children’s spelling samples, 36 (21.05%) of these were independently coded by two well-trained research assistants, and their interrater reliability was 92% (K = .92, p < .001). A systematic analysis of different types of errors involved in the present study across three categories of children’s spelling errors is shown in Table 4, separately for Times 1 and 2. Phonologically based spelling errors constituted approximately 3% and 5% of the total errors at Time 1 and 2, respectively. Of these, homophone and semi- TABLE 4 Distribution by Percentage and Number of Different Types of Spelling Errors, and Chi-Square Tests for Difference Between Time 1 and 2 for All Types of Spelling Errors in the Spelling Tests Time 1 Time 2 Error Category Error Type % (no.) % (no.) χ2(1, N = 171) Phonological Orthographic Morpho-lexical Total 1. Homophone 2. Semihomophonic Subtotal 3. Reconfiguration 4. Similar configuration 5. Stroke change Subtotal 6. Lexical 7. Sublexical 8. Single omission subtotal 2.0% (81) 0.8% (31) 2.8% (112) 0.6% (24) 3.0% (121) 2.1% (87) 5.7% (232) 0.7% (28) 1.3% (54) 89.5% (3,648) 91.5% (3,730) 4,074 3.8% (133) 1.3% (45) 5.1% (178) 1.2% (41) 12.0% (416) 9.1% (316) 22.3% (773) 1.2% (42) 4.3% (149) 67.1% (2,333) 72.6% (2,524) 3,475 22.37*** 4.84* 27.95*** 6.99** 228.60*** 178.29*** 443.66*** 5.00* 61.76*** 570.78*** 471.19*** *p < .05. **p < .01. ***p < .001. MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 441 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 homophone substitutions occupied around 2% and 1% of the Time 1 total errors, respectively, with percentages of 4% and 1% at Time 2. Homophone errors refer to using the homophone as a substitute for the character in the target word, for example, the character /zoi3/ (again) in the target word (see you again) was misspelled as its homophone form /zoi3/ (exist). Similarly, the semi- homophonic errors were characterized either as being substituted for the phonetic radical of the character, such as using the phonetic radical to represent /nai4/ (mud) in the target word (earth), or as writing a character with partially similar sounds as the character in the target word, for example, the character /dou6/ (truth) in the word (to know) was misspelled by children as /dou3/ (reach). One of the relatively salient types of errors was an orthographic-based error, consisting of approximately 6% and 22% of the total errors at Times 1 and 2, respectively. Orthographic-based errors included reconfiguration, similar configuration, and stroke change. These three types of error and their percentages were approximately 1%, 3%, and 2% of the total errors at Time 1, and 1%, 12%, and 9% of the total errors at Time 2, respectively. Reconfiguration refers to reversing the positions of the components of the character, for example, the character (mother) in the word (mother) wrongly spelled or switching the order of characters in the target words, for example, the target word (male female) was mis spelled by children as (female male). In the “similar configuration errors” children misspelled a character using a similar form and structure to another character, such as the character (know) in the target word (to know) being substituted (harmony). Stroke change errors focused on incorrect spellings at the stroke level, that is, the addition, subtraction, and protrusion of the strokes. For example, the first character (a man) in the target word (farmer) was incorrectly written as . The most predominant error was morpho-lexically based, accounting for 92% and 73% of the total error variance for Times 1 and 2, respectively. This category consisted of lexical errors. In these cases, children wrote a synonym, that is, a character with a semantic relationship to the target character. For example, the character /lim5/ (face) in the target word (wash face) was misspelled by children as /min6/ (face) or the character (flower) in (chrysanthemum) was written by children as (grass). A sublexical error was one in which children wrote the semantic radical or semantic-related part only. For example, the first character (jump) in the target word (dancing) was written by children as (foot) only. Finally, a single omission of the morpheme was when children left a single morpheme unwritten in a two-character word. Of all of these errors, lexical and sublexical types of mistakes constituted approximately 2% of the total errors separately at Time 1, whereas the single morpheme omissions accounted for almost 90% of the total errors. In other words, the largest error percentages were those of single morpheme omission. Similarly, lexical and 442 TONG ET AL. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 sublexical errors constituted almost 6% of the total errors at Time 2, and single omissions consisted of approximately 73% of the total errors. Apart from single morpheme omissions, we next tested which were the most prevalent errors of the spelling tests. A comparison of the frequencies of the remaining seven types of errors including homophone, semihomophone, reconfiguration, similar configuration, stroke change, lexical, and sublexical demonstrated that their frequencies were significantly different, χ2(6, N = 171) = 24.88, p < .001 at Time 2, but not at Time 1, χ2(6, N = 171) = 2.36, p > .05. A chi-square test was further used to examine the relations between time and spelling errors (phonological, orthographic, and morpho-lexical), and the relation was significant, χ2(2, N = 171) = 471.19, p < .001. This analysis indicated that children’s error frequencies changed with reading exposure. As shown in Table 3, children were more likely to make errors that could be categorized as homophone, semihomophone, reconfiguration, similar configuration, stroke change, lexical, and sublexical in the spelling test at Time 2 as compared to those in Time 1. However, omission of single morphemes decreased significantly as children’s reading knowledge increased. Finally, we analyzed the extent to which children’s spelling errors could explain children’s performances in Chinese character recognition, word dictation, and reading comprehension using hierarchical multiple regression analyses summarized in Tables 5 and 6 for Times 1 and 2, respectively. Every analysis included the vocabulary definitions task and age as control variables, and phonologically based errors, orthographic-based errors, and morphologically based errors were entered as a second step to investigate the contributions of each error type to the three literacy outcomes across time. As indicated in Table 5, these three types of spelling errors collectively explained approximately 28.5%, 72.4%, and 15.2% of the variance in Chinese character recognition, word dictation, and reading comprehension, respectively, at Time 1. The final standardized betas further revealed that phonologically based errors were uniquely and negatively associated with Time 1 Chinese word dictation but not with Chinese character recognition or reading comprehension. Orthographic-based errors were negatively associated with both Chinese word dictation and reading comprehension after controlling for age and vocabulary knowledge. In contrast, only the morphologically based error category was uniquely associated with all three literacy skills. Similarly, the analyses of these spelling errors on subsequent literacy skills also revealed that these three types of error explained almost 30.3%, 33.3%, and 14.2% of the variance in Chinese character recognition, word dictation, and reading comprehension when vocabulary knowledge and age were statistically controlled. Moreover, the final standardized betas further demonstrated that only the morphologically based errors were uniquely associated with each of these three literacy skills consistently. Across time, all three error types were uniquely associated with Chinese word dictation skills. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 TABLE 5 Hierarchical Regression Predicting Time 1 Chinese Character Recognition, Chinese Word Dictation, and Reading Comprehension From Three Types of Spelling Errors T1 Chinese CharacterT1 Chinese ReadingRecognition T1 Chinese Word Dictation Comprehension Steps and Predictors R2 R2 R2 R2 R2 R2 Vocabulary definitions, T1 1. .173 .173*** .227*** .134 .134*** –.022 .160 .160*** .127. Age, T1 –.053 –.009 .181* Phonologically based errors , T1 2. .457 .285*** –.100 .858 .724*** –.169*** .312 .152*** .004 Orthographic-based errors, T1 –.055 –.206*** –.182** Morphologically based errors, T1 –.604*** –.952*** –.412*** Note. T1 = Time1; T2 = Time2. . p < .10. *p < .05. **p < .01. ***p < .001. 444 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 TABLE 6 Hierarchical Regression Predicting Time 2 Chinese Character Recognition, Chinese Word Dictation and Reading Comprehension From Three Types of Spelling Errors T2 Chinese CharacterT2 Chinese ReadingRecognition T2 Chinese Word Dictation Comprehension Steps and Predictors R2 R2 R2 R2 R2 R2 Vocabulary definitions, T1 1. .159 .159*** .231*** .023 .023 .009 .136 .136*** .241 Age, T1 –.105 –.137 –.036 Phonologically based errors , T1 2. .462 .303*** –.050 .356 .333*** –.158* .278 .142*** –.115 Orthographic-base errors, T1 –.050 –.216** –.052 Morphologically based errors, T1 –.620*** –.619*** –.422*** Note. T1 =Time 1;T2 =Time 2. *p < .05. **p < .01. ***p < .001. MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 445 DISCUSSION Downloaded By: [University of Alberta] At: 08:44 19 December 2009 The present study examined a fairly wide variety of reading-related skills—that is, word recognition, word spelling, and reading comprehension—in a group of young Chinese children. We focused on cognitive components of literacy with a relatively comprehensive test battery. In the present study, morphological awareness and orthographic knowledge both appeared to be consistently associated with concurrent Chinese literacy skills across reading comprehension, word reading, and spelling even with other reading-related measures controlled, but phonological awareness did not. Morphological awareness was also longitudinally predictive of Chinese character recognition and reading comprehension, though not word dictation, whereas orthographic knowledge was only significantly related to subsequent word reading skills. Of interest, phonological awareness uniquely explained subsequent word dictation. Finally, our regression analyses demonstrated that speeded naming consistently accounted for unique variance of each of these three literacy skills beyond other variables, even with the autoregressive effects of each statistically controlled. The other relatively novel feature of the present study was the extensive spelling error analysis at the word level. This systematic analysis demonstrated the importance of spelling errors in explaining unique variance in concurrent and subsequent literacy skills. Although all three types of error analyses were important for some literacy skills either concurrently or longitudinally, those categorized as morphologically based were consistently associated with performance across literacy domains. The findings on morphological awareness extended previous research on the contribution of morphological awareness to literacy achievement in Mainland Chinese fifth and sixth graders with and without dyslexia (Shu et al., 2006), and on the roles of various cognitive-linguistic components involved in text reading comprehension among Hong Kong third through fifth graders (Leong et al., 2007; Leong et al., 2008) to demonstrate that morphological awareness is essential for understanding early Chinese literacy acquisition. Young Hong Kong Chinese children start to learn Chinese relatively early compared with other Chinese societies, and they are taught to focus on holistic word and character reading without the aid of phonetic transcription (e.g., Holm & Dodd, 1996). For exam ple, a two-character word (farmer) was learned by young children as an unanalyzed squared-shape image, directly mapped onto the corresponding speech sounds via rote memorization. Of interest, however, some children who could read the whole word could not read the individual characters when they were present individually, that is, (agriculture) or (man) by themselves. This focus on holistic processing may make morphological awareness and orthographic skills particularly important for learning to read and spell in Chinese for Hong Kong children. What is special about morphological processing for early literacy in Chinese? 446 TONG ET AL. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 The first plausible explanation for the importance of morphological awareness in young Chinese children’s literacy mastery is the salient semantic transparency of the compounding morphological structure in many Chinese words, facilitating understanding even of text passages from the contextual cues. Like some English morphologically complex words, a great number of Chinese words have meanings that can be inferred from their morphemes, for example, the meaning of the four-character compound word /mut6 lei6 faa1caa4/ (jasmine tea) may be predictable from the single morpheme /caa 4/ (tea). Children’s awareness of the lexical compounding structure of this and many other words might facilitate children to recognize, or correctly retrieve the single character in a multicharacter word, or even to understanding the gist of the meaning of a short comprehension passage. Moreover, such compound knowledge, which might promote flexibility in making use of context for reading sentences or longer passages, might even facilitate children to quickly extract the meaning of texts that include unknown words via an “educated guess.” As reported by Nation and Snowling (1998), context, when available, facilitates word reading in both typically developing and disabled readers. In addition, the large number of homophones of Chinese, in sharp contrast to the relatively few in alphabetic languages, makes phonological or sound information relatively unreliable in identifying or decoding characters. Moreover, children’s understanding of the homophonic nature of Chinese, in which identical sounds might represent different meanings in different word contexts, might enhance children’s accuracy in character naming and spelling. Homophones are not distinct from the aforementioned lexical compounding in Chinese, because the typicality of multimorpheme words for young children is probably attributable, in part, to the fact that such compounding helps to disambiguate these diverse meanings across homophones. The importance of morphological awareness is perhaps clearest at the word level, particularly in relation to word recognition, but it was also important for reading comprehension itself in the present study. However, as suggested by Nagy et al.( 2006), who talked about morphological awareness in relation to English reading comprehension, morphological awareness likely indirectly impacts reading comprehension by directly influencing vocabulary growth, in addition to its direct effects on reading comprehension via disambiguating text. Finally, another meaning-based aspect of Chinese words, that is, semantic radicals, either fully or partially cue meaning within most individual characters. Moreover, every individual character that includes the same semantic radical generally belongs to the same semantic category. Therefore, as children become more aware of the internal structure of print units via word specific learning (Chliounaki & Bryant, 2007), they may self-scaffold in reading or spelling. Some evidence supporting this claim is found in our spelling error pattern analysis, in which approximately 1% and 4% of errors, respectively, belonged to the sublexical level at Times 1 and 2. This indicates that Chinese children have a growing awareness of MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 447 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 the internal structure of Chinese characters, and they try to make use of relevant semantic radicals to represent meaning related to the target character. Perhaps partly related to this level of morphological processing, that is, at the subcharacter level, orthographic knowledge also appears to be a stable predictor of early Chinese literacy skills. Others have demonstrated that orthographic knowledge plays an important role in learning to read and spell Chinese (Ho et al., 2003; Shen & Bear, 2000) and in distinguishing dyslexic from nondyslexic readers (e.g., Ho et al., 2004; Ho et al., 2002). Given the strong emphasis on rote memorization of words in a holistic pattern with accompanying decomposition analysis in Hong Kong (Cheung & Ng, 2003), the importance of orthographic skills for these children is not surprising. At the same time, however, we have demonstrated the salience of these abilities in relatively young children and extended explorations of orthographic processing in Chinese literacy to reading comprehension itself. The types of spelling errors detected in our spelling analyses also underscored the importance of both orthographic knowledge and morphological awareness for Chinese children’s literacy development. Sound-based errors constituted a mere 3% and 5% (for Time 1 and Time 2, respectively) of the total number of errors, underlining the fact that phonological cues for Chinese word reading are relatively unreliable (Shu et al., 2003). Thus, correspondingly, it is not surprising that with other variables statistically controlled, the phonological awareness measure failed to explain unique variance across concurrent literacy skills. At the same time, most of the errors children made in the dictation task focused on orthographic and morphological confusions, rather than phonological errors. Most errors at the character level reflect children’s basic understanding of the forms and functions of Chinese radicals (e.g., Ho et al., 2003). However, these were most frequent at the morpheme level. This finding is in stark contrast to English or other alphabetic orthographies, where spelling is not usually centrally dictated by morpheme, at least at such a young age in children (but see Bryant, Nunes, & Bindman, 1999, for confusions in spelling past tense in English, e.g., ed vs. t). Chinese spelling errors are overwhelmingly likely to fall at the level of the morpheme, specifically such that a given character is unknown or forgotten. This pattern underscores the importance of morphological awareness for learning to read, because mapping spoken morphemes to written symbols is a central task of learning to read and write in Chinese. Although there was a tendency for these data to highlight the salience of morphologically based processing or orthographic skills for literacy development in the present study overall, we must also acknowledge the importance of reading-related skills that have been more of a traditional focus in the literacy on literacy development and impairment. In particular, apart from speeded naming, the only significant reading-related predictor of subsequent word dictation in the present study was phonological awareness. Given the prevalence of orthographically and mor 448 TONG ET AL. Downloaded By: [University of Alberta] At: 08:44 19 December 2009 phologically based errors in the spelling error analyses, this was somewhat surprising. On the other hand, phonological awareness is a universal correlate of early reading skills (e.g., Perfetti, Liu, & Tan, 2005; Ziegler & Goswami, 2005). Moreover, the phonological and morphological awareness measures were correlated at .54, perhaps underscoring how integrally related morphology and phonology, both represented at the syllable level, are in Chinese. The fact that speeded naming was uniquely associated longitudinally with all three literacy tasks in the present study, even controlling for the autoregressive effects of each one separately, is also noteworthy in the present study. Previous work (Ho & Lai, 1999; Shu et al., 2006) has identified speeded naming as a key correlate of reading variance in Chinese children. However, this work was virtually all focused on concurrent associations. In the present study, speeded naming emerged as particularly important for explaining literacy skills over time. None of the literacy tasks we tapped was timed, making this association all the more remarkable because part of the core of speeded naming skill is speed itself. The present study yet again underscores the potential importance of speeded naming for understanding the development of a wide range of reading-related skills in Chinese children’s development (e.g., Ho & Lai, 1999; Leong et al., 2008; McBride-Chang, Tong, et al., 2008; Shu et al., 2006). Because this was a correlational study only, we must be cautious in interpreting our findings. A variety of reading-related measures and reading outcome measures were administered to young Chinese children across the period of 1 year. From these data, it remains difficult to argue for causal associations of morphological awareness or orthographic processing and literacy skills per se, because, although they were sometimes uniquely associated with one or more of these abilities across the year, neither emerged as uniquely associated with any of the three reading-related tasks with the autoregressive effects of these tasks additionally statistically controlled. In addition, the instruction method of learning Chinese in Hong Kong tends to be quite different from those used in other Chinese societies, where phonological coding is emphasized more strongly and systematically (e.g., Cheung & Ng, 2003). Hence, whether the crucial role of morphological awareness in early literacy acquisition suggested in the present study can extend to other young Chinese children samples such as Mainland China, Singapore, or Taiwan requires additional testing across a wider range of Chinese societies. Despite these limitations, however, our results have at least two implications for early children’s literacy acquisition. First of all, both zero-order correlations and final standardized beta weights demonstrated that morphological awareness was fairly consistently associated with word reading, spelling, and reading comprehension. Moreover, spelling errors tended to be centered on the single morpheme level. These findings are consistent with previous studies from reading development and impairment (McBride-Chang et al., 2003; Shu et al., 2006) and suggest that at an early stage of reading development, children’s understanding of MORPHOLOGICAL AWARENESS AND ORTHOGRAPHIC KNOWLEDGE 449 Downloaded By: [University of Alberta] At: 08:44 19 December 2009 the morphological structure of their spoken language might be effective to help them to become good readers and writers later on. Second, although the pattern of spelling errors analyzed at the whole word level showed that children apply diverse strategies in spelling words, morpho-lexical and orthographic-based errors accounted for approximately 97% and 95% of the total errors across ages 6 and 7, respectively, suggesting that Hong Kong Chinese children tended to be relatively sensitive to the global picture of the two-character context of a word rather than the local features of a single given character. As an illustration of this point, some chil dren could correctly write the whole word (self), but they did not recognize the single c
相簿設定
標籤設定
相簿狀態