The Effectiveness of Diverse Designs of Chinese Stem-Deriving Instruction for Chinese Children With Dyslexia

Abstract

One of the most well-known instructional methods used to enhance Chinese character reading (CCR) is Chinese stem-deriving instruction (CSDI). In this method, CCR is taught via a group of characters (e.g., 清, 情, and 靚) that share the same stem (e.g., 青). However, the effectiveness of CSDI on the CCR of Chinese children with dyslexia remains inconsistent, and the reason may be the diverse designs of this method across studies. Thus, this study aimed to use a quasiexperiment-design study to compare the effects of various forms of CSDI based on the (in)consistency of the extending characters’ sounds, their appearances, and their amount (i.e., how many of them). In total, 122 Chinese third to sixth graders with dyslexia were recruited from primary schools, and they were randomly assigned to one of three groups: (a) visuospatially controlled CSDI (n = 37; all taught characters shared the same pronunciation), (b) verbally controlled CSDI (n = 45; all taught characters shared the same structure), and (c) noncontrolled CSDI (n = 40; all taught characters selected only on the basis of their frequencies). They were tested by CCR and dictation before and after the instruction. We found that CCR was significantly improved in all three CSDI groups, whereas only the verbally controlled CSDI group showed significant improvements in dictation. Furthermore, verbally and visuospatially controlled CSDI earned the greatest benefits in decreasing the visual errors and phonetic errors in CCR and decreasing the radical errors and homophone errors in dictation. Our findings shed light on the effectiveness of CSDI in terms of diverse designs, and frontline practitioners should take our findings into consideration when using CSDI.

Keywords

dyslexia Chinese stem-deriving instruction character reading dictation

Chinese character learning is inherently complex due to the unique characteristics of the Chinese writing system. Unlike alphabetic languages where letters represent phonemes, Chinese characters are logograms, where each character represents a morpheme and carries both phonetic and semantic information (Ho, Ng, et al., 2003). The characters are composed of strokes that form radicals, and the configuration of these radicals can vary greatly among characters. Moreover, a significant proportion of Chinese characters, known as compound characters, consist of a semantic radical that hints at the meaning and a phonetic radical that suggests the pronunciation (Shu et al., 2003). For instance, the semantic radical of 炮 (appearance of the Chinese character) (/pao4/ [its pronunciation in pinyin with a number indicating the lexical tone], gun [its meaning]) is 火 (/huo3/, fire), and its phonetic radical is 包 (/bao1/, bag). These unique features make Chinese character reading and writing a cognitively demanding task, requiring the coordination of various skills, including phonological awareness, orthographic knowledge, and morphological awareness (Tan & Perfetti, 1997). The idea of Chinese characters and the radicals can be found in the Character and Radical levels in Figure 1 (Yeh et al., 2017).

Figure 1.

This multilevel model summarizes four levels of Chinese characters that contain each Chinese character. From “Lexical Processing of Chinese Sub-Character Components: Semantic Activation of Phonetic Radicals as Revealed by the Stroop Effect” by S. Yeh, W. Chou, and P. Ho, 2017, Scientific Reports, 7(1), p. 2. Copyright 2017 by Springer Nature.

In this regard, although the prevalence of dyslexia in the Chinese language is estimated to be around 9.7% (Chan et al., 2007), which is not significantly different from that in alphabetic languages (Yang et al., 2022), the characteristics of Chinese dyslexia are distinct due to the uniqueness of the Chinese writing system.

Dyslexia, a major category within the spectrum of special learning needs, is considered a specific difficulty in learning to read and spell despite average intelligence and the absence of sensory and neurology limitations, emotional and behavioral disturbances, and environmental deprivation (Chung, 2016). Thus, providing effective instruction to support the reading and writing needs of children with dyslexia can be challenging for teachers, especially the needs above, and therefore alternative or specially designed approaches and methods are needed (Piotrowski & Reason, 2000). Mostly, these alternative or special designs are based on the idea of remediation of struggling reading performance as well as specific needs in cognitive or literacy abilities. For instance, researchers have found that phonological processing difficulties contribute to the reading challenges experienced by children with dyslexia (Ziegler & Goswami, 2005). In alphabetic languages, so many scholars have successfully proven that diverse designed phonological awareness–relevant training can have positive effects on the general reading performance of children with dyslexia, such as directly training individuals’ detection and manipulation of phonemes (e.g., Schneider et al., 1999), gamified phonological awareness training (e.g., Brennan et al., 2022), or using music as the tool to train individuals’ sensitivity (e.g., Flaugnacco et al., 2015). However, compared with the relatively consistent evidence in alphabetic languages, it is believed that Chinese children with dyslexia have more diverse needs than those with typical developments, including visual-orthographic knowledge, rapid naming, visual perception, and phonological awareness (Ho et al., 2002; Ho & Bryant, 1997). Among those, visual-orthographic skills, which were defined as the ability to recognize whether structure and components are correctly oriented (Badian, 2005), have been found to be an important factor in learning Chinese characters (Huang & Hanley, 1995), and a close connection between visual-orthographic information and sound was also reported (Ho & Bryant, 1997). Thus, many teaching methods designed to facilitate reading and writing Chinese characters in children with dyslexia, such as Chinese stem-deriving instruction (CSDI), have been generated by targeting this ability.

CSDI is designed based on reforming one of the key weaknesses of regular instruction: teaching new characters using an unsystematic approach because the choices of new characters must fit the sequence of teaching each text and in recognizing and reading Chinese characters in context (Wan, 1991). As proposed, in CSDI, Chinese characters are taught in a group with other characters that share the same stem. This is called stem-family instruction (Lu, 2000). For instance, the teaching design involves teaching the stem 包 (/bao1/, bag) and its extending characters with semantic radicals (i.e., the contained component represents the meaning of the compound Chinese character) such that 手 (/shou3/, hand), 火 (/huo3/, fire), and 草 (/tsao3/, grass) are 抱 (/bao4/, hug), 炮 (/pao4/, gun), and 苞 (/bao1/, bud) are thus taught within one teaching section.

On the basis of the previous evidence, CSDI is believed by some scholars and teachers to be an effective instructional methodology that can improve the learning reading and writing of Chinese characters for most Chinese children with dyslexia (Zhang et al., 2006). However, the difference between the effectiveness of CSDI and regular instruction in improving the character learning of children with dyslexia is debated. More specifically, some studies have indicated that children with dyslexia who learned characters via CSDI demonstrated significant learning of characters (Lu, 2000; Sun et al., 2006) and even outperformed those with dyslexia who learned via regular instruction (Chen et al., 2013). Contrary to this finding, other studies have reported that there was either no difference between the types of instruction or a slightly lower level of achievement for children with dyslexia who had received CSDI (Hu, 2001, 2005). These inconsistencies may be referred to the complexity of how much the shared radical of Chinese characters could affect students’ learning found by S.-J. Lin 1997. That is, in Lin's study, Chinese characters containing the same radical are more beneficial for poor-achieved students, whereas average- and well-achieved students had better performance in dictating those Chinese characters not containing the same radical.

Such discrepancy may be due to the designs of the CSDI itself (for a review, see C.-C. Wang, 2005). Stem and extending characters are two main components of stem-deriving instruction. The stem chosen tends to be consistent across studies because most of the studies considered the high-fluency stems for primary school–age Chinese students to be a benchmark. However, choosing extending characters is much more inconsistent across relevant studies than choosing stems because neither most of the previous studies’ research designs nor practitioners’ teaching designs had a clear principle in choosing extending characters. C.-C. Wang (2005) further argued that most stems involved in this instruction are phonetic radicals, so children who receive a group of Chinese characters with similar or the same sounds may face a certain degree of difficulties to follow, especially for those with lower levels in phonological processing. It was argued that such challenges were because lower levels of or slower phonological processing could impede discriminating several stimuli with the same or similar sounds. Additionally, a group of Chinese characters with very similar appearances could increase learning difficulties, which could raise the burden of distinguishing several characters presented simultaneously, especially for those with lower levels of visual perception. It is noteworthy that the heterogeneous nature of Chinese students with dyslexia, including that some of them mainly have needs in the detecting and manipulating of auditory information (i.e., phonological processing) whereas some mainly have needs in recognizing and structuring components of Chinese characters (i.e., orthographic knowledge), was found in the previous studies (L.-C. Wang & Yang, 2014). Thus, it is necessary to consider adapting the details of stem-deriving instruction, such as adjusting the similarity of pronunciations or appearances of those chosen extending characters, to benefit as many students as possible.

From this perspective, even though it is common to select those highly frequent Chinese characters as the extending characters, three factors may influence could which extending characters are selected, that is, their sounds, their appearances, and their amount (i.e., how many of them) (e.g., Chen et al., 2013; Hu, 2001, 2005; Lu, 2000; Zhang et al., 2006). Thus, it is reasonable to infer the necessity of providing proper numbers or complexities, in terms of sounds or appearance, of materials to better fit the cognitive loadings of Chinese students with dyslexia in each of the aforementioned aspects. Furthermore, the cognitive loadings in these aspects could correspond to different working memory domains, mainly including verbal (phonological loop) and visuospatial (visuospatial sketch pad) domains (Baddeley, 1986).

Although such increased working memory loads may not put typically developing children's learning in jeopardy, it is highly possible to affect the learning of Chinese children with dyslexia because the particular need for working memory in this group has been preliminarily identified the needs in working memory in Chinese children with dyslexia (e.g., Ho et al., 2004; Y. Luo et al., 2013). Challenges with working memory in different domains are believed to be barriers to domain-specific skills, knowledge, and procedures (Peng et al., 2015).

According to cognitive load theory, to facilitate learning, instructional methodologies should attempt to eliminate, to the greatest degree possible, working memory demands (e.g., Rougier & Bonnet, 2016). However, when selecting the extending characters for CSDI, the working memory loads of different domains may increase according to their frequency or level of difficulty, which may lead to controversy regarding CSDI’s effectiveness, especially for those with dyslexia who are already experiencing limited working memory. Thus, to enhance the effectiveness of CSDI for children with dyslexia, the clear relationships between and among working memory and the various aspects of extending characters must be understood.

Verbal Working Memory: Pronunciation of Extending Characters

Unlike English's assembled phonology involving letter-sound mapping, the pronunciation of a Chinese character is derived from its phonetic radical or by analogy with another character sharing the same radical, a feature known as orthography-phonology correspondence (OPC) rules. These apply to approximately 80% of Chinese characters, the compound ones, with the remaining 20% being single characters with no OPC. Chinese compound characters are categorized based on phonetic regularity into regular, semiregular, and irregular characters. Regular characters, like 清 (/ching1/, clear), have a phonetic radical identical in pronunciation to the whole character, making up 23% of characters in primary school textbooks. Irregular characters, such as 法 (/fa3/, law), where the radical pronunciation differs from the whole character's, comprise 15% of characters. Semiregular characters, with a more flexible definition, account for 42% of characters and are divided into on-set, rime, and tone-difference categories based on variations in sound components. Phonetic regularity rules, similar to learning alphabetic scripts, are crucial for children learning to read Chinese characters, aiding in the understanding of phonetic radicals and lexical tones. However, dealing with a high variety of sounds can challenge the verbal working memory, especially in students with dyslexia. Consistency of phonetics, like choosing extending characters with identical pronunciations for each stem, has been shown to lessen the load on verbal working memory and improve the effects of CSDI.

The verbally controlled condition, where characters share the same sound, might indeed introduce a similarity effect that could potentially overload working memory. According to the working memory theory (Baddeley, 1986), if too much similar auditory information is presented simultaneously, it could exceed the capacity of the phonological loop, leading to a decline in processing efficiency. However, the studies (e.g., Wong et al., 2007) found improvements in reading and dictation in this condition, suggesting that the similarity in sounds may also have a facilitative effect on learning, potentially due to the reduced cognitive load associated with remembering different pronunciations. This intriguing finding calls for further investigation.

Visuospatial Working Memory: Structures of Extending Characters

Chinese orthographic knowledge, which denotes the specific placement of semantic and phonetic radicals in characters, is crucial for children learning to read and write. This knowledge system reduces processing units from over 10,000 characters to around 1,000 radicals. According to Ho, Ng, et al.'s (2003) model, visual-orthographic knowledge of Chinese characters develops in two stages: structural and positional knowledge. The former pertains to rules governing the combination of radicals into characters and aids in character separation. Examples of structural knowledge include 11 types of configurations, such as vertical and horizontal combinations, with vertical combinations like left-right and left-middle-right accounting for 47.04% and horizontal combinations taking up 33.87%. The remaining nine types share the remaining 19.09% frequency. Positional knowledge, the more complex stage, refers to the legitimate placement of phonetic and semantic radicals within characters, expanding the number of recognizable characters. Children's positional knowledge development varies, with primary third graders demonstrating 97% accuracy for semantic radicals and primary fifth graders achieving 82% accuracy for phonetic radicals. A simultaneous focus on diverse Chinese character structures may burden the visuospatial working memory, potentially making it difficult for readers, especially those with dyslexia, to learn these characters. Thus, offering corresponding structures of extending characters for the CSDI is considered an effective approach to alleviate the visual working memory load for dyslexic children and improve CSDI's efficacy.

Thus, in the visually controlled condition, where characters share different sounds, the variation in auditory information could indeed contribute to the working memory load. According to cognitive load theory (Sweller, 1988), learning is hindered when the working memory load exceeds the learner's cognitive capacity. The constant switch between different sounds in this condition could increase the intrinsic cognitive load, making it harder for learners to process and retain the information. However, this condition also showed improvements, suggesting that the diversity in sounds might have provided richer phonetic cues, helping learners to differentiate and remember the characters.

It is noteworthy that a growing body of evidence has revealed that instructional methods with diverse designs for reading and writing could possibly lead to different error patterns (e.g., Elgün-Gündüz et al., 2012; Rahimi, 2021). Thus, it is reasonable to infer that students with dyslexia receiving diverse designed CSDI may show, in addition to their overall performance in reading and writing, various error patterns in their Chinese characters or dictation. Such observations of error patterns are more useful for spotting students’ difficulties during their reading and writing.

In summary, the varying capacities of working memory of children with dyslexia in various domains may confine their Chinese character learning if the teaching designs are not appropriate. Therefore, this study intends to examine the effects of various types of CSDI based on the considerations of working memory loading on diverse modalities. More specifically, there are three types of instruction in this study: (a) verbally controlled CSDI, (b) visuospatially controlled CSDI, and (c) noncontrolled CSDI.

Research Aims

Accordingly, the research aims of this study are twofold. First, we aimed to compare each group's improvements in reading and writing performance after the instruction, and the intergroup differences in groups’ posttest performance compared with their pretest performance were also examined. Also, we compared the error patterns of Chinese character reading and dictation after the instruction in the three groups.

Method

Participants

In total, 122 Chinese children with dyslexia were recruited in primary schools from the third to sixth grades in Taiwan. In the beginning, 135 students joined this project; however, 13 of them withdrew due to various issues, primarily related to parental concerns. In this study, all eligible children with dyslexia had a standard score of −1.5 standard deviations below average on the Chinese word reading test and exhibited normal intelligence based on an IQ score greater than 80 on the nonverbal IQ test. These criteria for dyslexia in the present study are consistent with those of previous studies (Lyon et al., 2003). The pool of potential participants was officially diagnosed with dyslexia or learning disabilities in the reading area by the local authorities. Specifically, a committee of identification organized by a group of professionals, including professors, medical doctors, psychologists, and so on, is in charge of the identification in the special education system in Taiwan. We approached those students via the special education network, and the special education teachers’ and parents’ consent was obtained for further proceeding with the screening procedure. Additionally, all the selected participants should have normal or corrected-to-normal vision and hearing and no diagnosed attention deficit hyperactivity disorder.

The participants were randomly assigned to one of three groups, namely, (a) a visuospatially controlled CSDI group (n = 37; male = 20), (b) a verbally controlled CSDI group (n = 45; male = 24), and (c) a noncontrolled CSDI group (n = 40; male = 19). The chi-square test showed gender did not have significant linear association with group, χ²(1, N = 122) = 0.41, p = .56, so women were more likely than men to be involved in any group. The detailed differences of the three CSDI groups will be introduced in the Procedures section. The number of participants in each group at the beginning of the current study was 45, and the attrition rate was 17%, 0%, and 11% for each group, respectively. The main reasons for the participants’ attrition included time clashes with their after-class activities, feeling tired, and not being interested.

Measures

The participants were given a number of cognitive and literacy tests, including standardized tests and self-developed tasks. The participants were tested individually or within their group during one or two sessions for each testing period.

Raven's Standard Progressive Matrices

Nonverbal IQ, tested by Raven's Standard Progressive Matrices, has been considered a crucial index of the reading acquisition of children with and without dyslexia across studies (e.g., Chung et al., 2008; Wang & Yang, 2018). This is a standardized nonverbal IQ test that consists of 60 items of increasing difficulty. Each item has a target visual matrix with one missing part. The children are provided with six to eight response options and are asked to select the response that completes the missing piece in the visual matrix (Raven, 1996). This test was used to screen participants for the study.

The Chinese Character Recognition Scale

This test, developed by Huang (2001), is a standardized test in which participants read aloud the pronunciations of Chinese characters visually presented in a list. During the test, the participants were tested individually and were asked to utter the sounds as quickly as possible; the examiners could not provide cues or hints to the participants. Additionally, the examiners were required to write down the participants’ response, if it was incorrect, by using the phonetic symbol system in Taiwan. The test arrangement includes 20 lines, and each line includes 10 Chinese characters, totaling 200 Chinese characters. The applicable targets of this test are first to ninth graders. Considering the large sample size in this study, the participants were given a 7-min time limitation. Based on our sample and procedures, Cronbach's α coefficient was .75.

The Battery of the Chinese of Pupils (Dictation Subtest)

This standardized battery of tests was developed by Hung et al. (2003) to comprehensively assess Chinese children's reading and writing skills at the character level. To examine the purpose of the present study, only dictation was used. In this subtest, the participants were requested to write the target characters they heard from the examiner's indication in a two-character vocabulary and one character for each item. There are 45 characters contained in the Dictation subtest, and participants are requested to accomplish this subtest in 2 min. The average Cronbach's α coefﬁcient is .87 (p < .01), and the split-half reliability of this measure is .90 (p < .01) according to the handbook. This measure was used to assess participants’ literacy performance before and after the instruction.

Procedure

This is a quasiexperimental study in which the differences among the three groups of CSDI were in the consistency of the extending characters’ structures and pronunciations of the extending characters but not teaching approaches or steps. In other words, the instructional time, the number and schedule of instructional sessions, and the instruction providers’ strategy used across the three groups were identical.

The main teaching materials for CSDI are stems and extending characters, and they were selected in terms of the following principles. First, the stems of CSDI were selected from the textbooks of first through third grades and then analyzed. Thus, the participants were expected to be familiar with the stems. Second, the extending characters were selected by the degree of frequency and level of difficulty. The characteristics of extending characters for the three types of CSDI were selected to meet the participating students’ current status of (un)known Chinese characters. In most cases, we selected Chinese characters that are at approximately the first- to third-grade difficulty level.

There are two types of CSDI designed for working memory loadings for this study. The first is the verbally controlled CSDI, designed to reduce the learner's verbal working memory load and thus reduce the extra cognitive loadings caused by learning the target characters with varied sounds. In this group, the pronunciations of the extending characters were controlled and considered to have very high or identical consistency, whereas the structure of the extending characters was not consistent. In contrast, the visuospatial-controlled CSDI was designed to reduce the learner's visuospatial working memory load, thereby reducing the extra cognitive loadings caused by learning target characters with varied sounds. In this group, the structure of extending characters was controlled as having complete consistency, whereas the pronunciations of the extending characters were not consistent in the visuospatial-controlled CSDI. Furthermore, a type of CSDI was designed without considering working memory loadings (i.e., noncontrolled CSDI). For this type of CSDI, the structures and pronunciations of the extending characters are not consistent.

As mentioned before, the participants in the present study were divided into three groups, and their major difference is the extending characters presented during instruction. For the verbally controlled group, the extending characters’ pronunciations were controlled as having very high or complete consistency, whereas the structures of extending characters were not consistent. For the visuospatially controlled group, the structure of the extending characters was controlled for identical consistency, whereas the extending characters’ pronunciations were not consistent. For the noncontrolled group, the pronunciations and structures of the extending characters were inconsistent, but the number of extending characters was larger than that for the other two groups. In addition to the difference in the selected Chinese characters among the three groups, all the instructional designs and procedures were identical.

In order to better confirm the fidelity of the instructions in this study, the teaching for the three groups all followed the same approach of CSDI, and it is based on the original idea of introducing stem and then teaching extending characters (Wu, 2007). The detailed steps of this instruction are adapted and amended from Lu, 2000 teaching procedure with a combination of Wu, 2007 teaching designs for radical group text teaching strategies in which the contextual information of each Chinese character is emphasized on top of the basic idea of CSDI. The teaching flows of each session in all groups are identical, and the details are shown in Online Appendix A.

Furthermore, to facilitate the participants’ learning motivations and outcomes, all selected teaching materials included not only Chinese characters and radicals but also relevant multiple-media materials (i.e., graphics and animations). Furthermore, to strengthen participant memory, all multiple-media materials of the stems and extending characters were presented by referring to their sounds, meanings, and/or sources. An example of the teaching slides (except “Warm-Up Exercise” and “Assessment of Extension Activities” in Online Appendix A) appears in Online Appendix B.

In all CSDI groups, there were three to five participants in a small group receiving instructions in which each teaching session contained two or three extending characters. They received 50 min of instruction per session, and there were two teaching sessions per week for a total of 20 teaching sessions over a 12- to 14-week period. In each teaching session, as indicated earlier in the teaching steps, one Chinese stem was used as core knowledge, and varied extending Chinese characters were introduced. In the three groups, the Chinese stem used in every session was the same, but the selected extending characters were different by referring to the principles mentioned earlier. The details of the selected extending characters are in Online Appendix C. As shown in Online Appendix C, we were able to select only 10 stems due to the balance of all conditions, but in total 20 teaching sessions were held during the instruction phase. Thus, in the present study, each stem was taught twice.

Additionally, we invited two or three instructional providers as a group, and there was a total of three groups in the present study. They were all registered teachers in Taiwan. Because of the possible impact of the instructors, the matching of instruction providers and the small groups of the three CSDI groups was counterbalanced on the basis of teaching sessions. The instruction providers taught different sessions for every small group containing three to five participants. Also, although those instruction providers were expected to have some prior knowledge in teaching, they were also trained by the authors by (a) introducing the theoretical background of CSDI, (b) elaborating the details of this instruction in this study, and (c) asking them to demonstrate their teaching skills of this instruction.

Data Analysis

To examine the research aims of the present study, multiple statistical methods were implemented. First, we implemented a one-way ANOVA to conduct the pretest equivalence analyses. Then we introduced two sets of hierarchical linear models (HLMs), with study measures (i.e., Chinese character reading and dictation) as dependent variables; group, assessment time point, and their interaction as fixed effects; and small group as random effects, which means the influence of receiving the instructions in different small groups was considered to vary randomly, to test the first research aim. Also, to test the effectiveness of different designed CSDI on Chinese character reading and dictation, participants’ age, gender, and nonverbal IQ were controlled.

Second, we aimed to look into the details of the participants’ error patterns in their Chinese character reading and dictation. In this regard, two sets of one-way MANCOVAs were used to compare the error types on posttest Chinese character reading and dictation separately while controlling for age, gender (coded as 1 and 2 as a variable in the statistical analysis), and nonverbal IQ.

More specifically, six error patterns for reading Chinese characters and for vocabularies were observed according to Shu et al. (2005) and L.-C. Wang and Yang (2014), including (a) semantic errors (naming the target character as another character with similar meaning), (b) selective errors (naming the target character as a character that forms a highly frequent two-character compound word), (c) visual errors (confusing the target character with a visually similar character), (d) phonetic errors (naming an irregular character after the phonetic radical), (e) analogy errors (naming a character after another character with the same phonetic radicals when such pronunciation is incorrect), and (f) homophone errors (using another character with the same pronunciation as the target character to form words with the meaning character). Another five error patterns in writing Chinese characters or vocabularies were observed by adapting scales from Law and Or (2001) and Law et al. (1998), including (a) stroke errors (stroke reversal; addition, substitution, and omission of stroke), (b) radical errors (broken stroke of a component, slanted structure, rectangular structure, and maze structure), (c) homophone errors (using a homophonic character to replace the target character), (d) phonetically similar errors (including similar onset, rime, and tone), and (e) unrelated errors.

To ensure the reliability of error pattern identification in our study, two certified special education teachers were invited to make judgments of the error patterns of each character, and the interrater reliability reached .93 and .81 in character reading and dictation, respectively, which were tested by using 20 incorrect reading and 20 incorrect writing performances (not from the data in the present study) beforehand. For the errors in character reading, judgments were based on the participants’ incorrect responses on the Chinese Character Recognition Scale. At the same time, raters judged the error in dictation and the participants’ inaccurate performance in the Battery of the Chinese of Pupils (Dictation subtest). It is noteworthy that it is possible to observe more than one type of error occurring for each character, whether on the Chinese character reading test or dictation test.

Results

Before examining the first research aim, pretest equivalence analyses were tested by conducting a one-way ANOVA to ensure the differences across the three groups were insignificant. The results, as shown in Table 1, indicated none of the background information (including gender, which was tested using chi-square in the Participants section) and pretest performance in Chinese character reading and dictation of the three groups had overall significant differences. After confirming the equivalence of the three groups, we further tested the effectiveness of differently designed CSDI on Chinese character reading and dictation; two sets of HLMs were used. The results are integrated in Table 2.

Table 1.

Pretest Equivalence Analyses of Three Groups.

	VisCon(n = 37)		VerCon(n = 45)		NC(n = 40)		F(2, 119)	Significance
Variable	M	SD	M	SD	M	SD	F(2, 119)	Significance
Age	9.99	0.97	9.88	0.92	10.21	0.98	1.27	0.29
Nonverbal IQ	27.59	5.61	27.22	5.44	28.30	5.52	0.41	0.66
CCRPre	53.08	19.26	51.78	17.92	56.20	19.63	0.60	0.55
DICPre	14.84	4.64	14.38	3.79	15.35	3.87	0.60	0.55

Note. VisCon = visuospatially controlled Chinese stem-deriving instruction (CSDI); VerCon = verbally controlled CSDI; NC = noncontrolled CSDI; CCRPre = pretest Chinese character reading performance; DICPre = pretest dictation performance.

Table 2.

Final Fixed-Effects Model Outcomes and Variance Components for Chinese Character Reading and Dictation, Controlling for Gender, Age, and Nonverbal IQ.

	Chinese character reading				Dictation
	Estimate	SE	Statistic	p	Estimate	SE	Statistic	p
Predictor
Group [2]	−2.00	3.94	−0.51	.612	−0.43	1.04	−0.42	.676
Group [3]	3.48	4.05	0.86	.392	0.46	1.07	0.43	.669
Time [2]	30.98	4.62	6.70	<.001	1.89	1.09	1.74	.084
Age	0.77	1.06	0.72	.470	0.02	0.32	0.05	.959
Nonverbal IQ	−0.05	0.18	−0.27	.791	0.06	0.06	1.14	.255
Gender [2]	−6.42	2.00	−3.21	.001	0.06	0.61	0.10	.919
Group [2] × Time [2]	2.33	6.21	0.38	.708	14.35	1.47	9.76	<.001
Group [3] × Time [2]	0.56	6.38	0.09	.930	−0.84	1.51	−0.56	.578
Random effects
σ²	199.86				21.94
τ₀₀	30.62_SmallGroup				0.00_SmallGroup
τ₁₁	104.12_{SmallGroupTime2}				0.00_{SmallGroupTime2}
Marginal R² / Conditional R²	0.504 / 0.631				0.617 / NA

Note. Numbers in brackets refer to the number of variables.

The significant interaction between timing (pretest and posttest) and group was found only on dictation and not on Chinese character reading. Thus, for Chinese character reading, the main effects were calculated for the difference in the performance between two testing times (pre- and posttest). The results indicated that the posttest performances in Chinese character reading of all the participants were significantly better than their pretest performances (p = .00). In contrast, without separating two testing times (pre- and posttest) of performance, a nonsignificant difference was found in every paired comparison (i.e., visuospatially controlled CSDI group vs. noncontrolled CSDI group, visuospatially controlled CSDI group vs. verbally controlled CSDI group, and verbally controlled CSDI group vs. noncontrolled CSDI group) in Chinese character reading.

Regarding the random effects, the intercept at the small-group level had a variance component of τ₀₀ = 30.62. For small-group Time 2, there was a notable variance component estimated at τ₁₁ = 104.12. The estimated covariance between the small-group-level intercept and small-group Time 2, denoted as ρ_01, was −0.25. Additionally, the variance within each small group, also termed as the residual variance, stood at σ² = 199.86.

Differing from Chinese character reading, students’ dictation showed a distinct profile in this study. Because there is a significant interaction on dictation, the simple main effects should be examined separately. The detailed results showed that there was no significant difference between pretest and posttest performance in dictation of visuospatially controlled CSDI (p = .07) and noncontrolled CSDI (p = .23). In contrast to these two groups, the posttest dictation performance of the participants in the verbally controlled CSDI condition showed significant improvement over their pretest performance (p = .00). Additionally, from the view of group differences in the posttest performance, the overall difference in dictation was significant (p = .00). Specifically, the verbally controlled CSDI group outperformed the other two groups (compared with the visuospatially controlled group, p = .00; compared with the noncontrolled group, p = .00), whereas these two groups (i.e., visuospatially controlled CSDI and noncontrolled CSDI groups) had no significant difference (p = .68).

For the random effect, the variance component for the intercept (small-group level) was not significant, with an estimate of τ₀₀ = 0.00. Similarly, the variance component for small-group Time 2 was not significant, with an estimate of τ₁₁ = 0.00. The covariance between the intercept and small-group Time 2, ρ₀₁, was −1.00. The within-small-group variance (residual variance) was σ² = 21.94.

To examine the effectiveness of different designed CSDI in students’ performances in Chinese character reading and dictation, as the second research aim, we used two sets of MANCOVAs. The error patterns of Chinese character reading and dictation as well as the comparisons are shown in Table 3. It is noteworthy that Bonferroni corrections were implemented to avoid Type I errors because many comparisons were conducted simultaneously.

Table 3.

Comparisons of the Error Types on Chinese Character Reading and Dictation of Three Groups’ Posttest Performance, Controlling for Age, Gender, Nonverbal IQ, and Small Group.

Error pattern	VisCon(n = 37)		VerCon (n = 45)		NC(n = 40)		F(2, 116)	Post hoc	Hedges's g (1v2)	Hedges's g (1v3)	Hedges's g (2v3)
Error pattern	M	SD	M	SD	M	SD	F(2, 116)	Post hoc	Hedges's g (1v2)	Hedges's g (1v3)	Hedges's g (2v3)
Reading
Semantic errors	7.73	0.8	7.53	0.94	7.58	0.64	0.69	1 = 2 = 3	−0.26	−0.29	0.07
Selective error	6.84	1.01	6.93	1.01	7	1.09	0.32	1 = 2 = 3	0.09	0.14	0.06
Visual errors	10.78	0.58	9.27	0.78	10.8	0.79	34.74*	2 < 1 = 3	−3.08	0.04	2.46
Phonetic errors	10.73	1.24	11.82	1.34	11.78	1.33	3.66*	1 < 3	0.64	0.63	−0.02
Analogy errors	9.51	1.12	9.33	0.77	9.28	0.45	0.38	1 = 2 = 3	−0.20	−0.32	−0.12
Unclassified errors	4.49	1.12	4.56	1.24	4.6	1.06	0.19	1 = 2 = 3	0.05	0.09	0.03
Dictation
Stroke errors	4.08	1.19	4.47	1.65	4.6	1.78	0.71	1 = 2 = 3	0.18	0.22	0.04
Radical errors	7.86	1.13	6.87	0.92	8.93	1.21	28.55*	1 = 2 < 3	−0.94	0.77	1.80
Homophone errors	4.32	1	5.58	1.12	5.65	1.12	11.92*	1 < 2 = 3	1.09	1.16	0.06
Phonetically similar errors	3.08	1.19	3.47	1.65	3.6	1.78	0.71	1 = 2 = 3	0.18	0.22	0.04
Unrelated errors	0.81	1.47	0.96	1.58	0.78	1.42	0.33	1 = 2 = 3	0.06	−0.01	−0.08

Note. VerCon = verbally controlled Chinese stem-deriving instruction (CSDI); VisCon = visuospatially controlled CSDI; NC = noncontrolled CSDI.

p < .008 or .01. Due to the application of the Bonferroni correction, the significance level for error types within the Reading domain has been set at p < .008, whereas for those in the Dictation domain, the threshold is p < .01.

As shown in Table 3, the error patterns of Chinese character reading and dictation are similar but different. Specifically, among all six error patterns of Chinese character reading, the three groups’ posttest performances showed significant differences only in the numbers of visual errors (confusing the target character with a visually similar character) and phonetic errors (naming an irregular character after the phonetic radical). Regarding the visual errors, the group receiving verbally controlled CSDI made significantly fewer numbers than the other two groups (comparing with the visuospatially controlled group, p = .00; comparing with noncontrolled group, p = .00), which did not perform differently in statistical meaning (p = 1.00). In contrast, students who received visuospatially controlled CSDI made fewer phonetic errors than noncontrolled CSDI (p = .00), whereas these two groups had no significant difference compared with verbally controlled CSDI in this type of error pattern (visuospatially controlled CSDI, p = .56; noncontrolled CSDI, p = 1.00).

On the other hand, the effectiveness of different designed CSDIs, at the statistical meaning level, could be found in radical errors (broken stroke of a component, slanted structure, rectangular structure, and maze structure) and homophone errors (using a homophonic character to replace the target character). For radical errors, students in the visuospatially controlled group (p = .00) and verbally controlled CSDI group (p = .00) made significantly fewer errors than the noncontrolled group, whereas the former two groups made no significantly different amount of this type of error (p = .08). Additionally, those who received the visuospatially controlled CSDI made significantly fewer homophone errors than those in the other two groups (compared with the verbally controlled group, p = .00; compared with the noncontrolled group, p = .00), which did not perform differently in terms of statistical meaning (p = 1.00).

In summary, in terms of our results, differently designed CSDI could lead to different effects in Chinese character reading and dictation in terms of both average performance and detailed error patterns. All three types of CSDI could clearly significantly improve the students’ Chinese character reading, but only receiving verbally controlled CSDI could help the students have significant increases in their dictation performance. Furthermore, verbally controlled CSDI has the advantages of lessening students’ visual errors, whereas visuospatially controlled CSDI's advantages are more obvious in decreasing phonetic errors. Finally, the unique strength of receiving visuospatially controlled CSDI is to reduce students’ homophone errors, and its function for decreasing radical errors is also strong; however, it is still weaker than receiving verbally controlled CSDI.

Discussion

This study examined the effects of various types of CSDI based on the considerations of working memory loading on diverse modalities. The results from the current study could be used to reflect the proposed research aims.

First, we compared the intragroup differences in the performances of the groups after the posttest and pretest. Our results revealed that regardless of the kind of CSDI the students received, their Chinese character reading significantly improved after 20 teaching sessions, and such improvements were basically no different among differently designed CSDI. This result reflects the general understanding of the effectiveness of CSDI in the past (e.g., Lu, 2000; Zhang et al., 2006). Also, it demonstrates the general understanding that giving a short-term instructional method in which Chinese characters with the same stem are taught in a group could be beneficial for Chinese character reading for those with dyslexia (Lu, 2000).

In contrast, the significant effect of CSDI on dictation was found only for those who received verbally controlled stimuli in our study. This finding could reflect the somewhat mixed previous findings regarding CSDI's effects on dictation (Lu, 2000; Hu, Hu, 2001, 2005). Such complexity mainly comes from the uncategorized designs of CSDI—the composite emphasis in the designs of the extended characters. Additionally, it is reasonable to explain our results by referring to reading-writing research. Basically, reading and writing are two skills that rely on similar underlying mechanisms (Abbott et al., 2010), so the transferring effect from reading to writing could be reasonably expected (Ehri, 1997).

However, the gap between word reading and writing may be due to many reasons, and the difference between phonetic retrieval (retrieving words’ sounds from seeing words) and orthography retrieval (retrieving words’ appearances from hearing sounds) is the most dominant. Although the natures of phonetic retrieval and orthographic retrieval are to retrieve information from the lexicon (Vitevitch, 2008), the former is much more difficult. This may be because the orthography retrieval in the current study required the participants, in addition to sound matching and retrieval, to use many other skills. For instance, they needed to rely on intact quantity of their mastered vocabularie and good morphological awareness to identify the right character they should write down. It is particularly difficult for Chinese readers because there are many homophones at both the character level (e.g., /shou4/could be 受, “by”; 授, “grant”; or 售, “sell”) and the vocabulary level (e.g., /cheng2 xin1/ could mean 誠心, “sincerity,” or 成心,“deliberately”). However, in the dictation task, to avoid the confusing situation, homophones at the vocabulary level were intentionally skipped, and the participants still needed to choose the right target character from approximately 11 choices (on average, 11 Chinese characters have the same syllable; Tan & Perfetti, 1997). Additionally, even eliminating the nature of Chinese and the limitation of the dictation task, it is also evident that people tend to perform more accurately in letter-to-sound-like tasks (closer to phonetic retrieval) than in sound-to-letter-like tasks (closer to orthography retrieval) (e.g., Meng et al., 1996).

Given the strengths verbally controlled CSDI has—releasing students’ cognitive energy from multiple pronunciations—it is unsurprising to learn that students in this group could better master how to retrieve the correct appearances of Chinese characters than those in the other groups, and that is orthographic knowledge. Orthographic knowledge enables remembering the visual and spelling patterns of individual words or word parts when seen on a printed page or, more simply, the conventions used in the writing system (Barker et al., 1992), and being sensitive enough to the correct orthographic structure of Chinese characters is one of the key features of this skill. In terms of Ho, Ng, et al.'s (2003) model, most of the children would have developed sufficient Chinese character structures before 5 years old, and such sensitivity to the structures of Chinese characters has been demonstrated as a unique factor independent of other features of orthographic knowledge (C. Luo et al., 2017). The findings regarding the lower sensitivity to the structures of the Chinese characters of Chinese children with dyslexia are reasonable (Chung et al., 2012); thus, lower levels of orthographic knowledge are almost the least debated core challenge of Chinese children with dyslexia among “multiple deficit hypotheses” (Ho et al., 2002). In this regard, the emphasis on Chinese characters’ structure in every instruction in this group of CSDI could no doubt free the students from spending more effort memorizing the various pronunciations of Chinese characters taught in one session and, in turn, lessen their cognitive loadings in becoming familiar with different structures and all other focuses at the same time.

Second, we compared the error patterns of Chinese character reading and dictation after the instruction of three groups and found clearly distinct advantages from differently designed CSDI. That is, those who received verbally controlled CSDI showed fewer visual errors in character reading and fewer radical errors in dictation. In comparison, those in the visuospatially controlled CSDI performed fewer phonetic errors in character reading and fewer homophone errors in dictation.

Generally, our findings regarding the error patterns of character reading and dictation further support the strengths of differently designed CSDIs. As we discussed earlier, those who received verbally controlled CSDI seem to have improved mostly in their structures of Chinese characters, and this skill with sensitivity to the orthographic structure is the developmental foundation of the following orthographic knowledge (Ho, Ng, et al., 2003). Among those, positional knowledge, that is, knowing the position of each radical in the Chinese script (D. Lin et al., 2019), could be considered the more sophisticated version of sensitivity to Chinese characters’ structures. In this regard, it is reasonable to infer that the students receiving verbally controlled CSDI may also help the students know the suitable positions of various radicals precisely, so it could prevent them from confusing some similar-looking characters containing similar radicals at other positions. Furthermore, although the students in the visuospatially controlled CSDI learned only one kind of Chinese character structure at a time, the structures they faced in different teaching sessions varied. Such variety is more systematic than in the noncontrolled CSDI group, because in the noncontrolled group there are two factors, that is, structures and pronunciations, that change together. This may be the reason that students in the visuospatially controlled CSDI group also showed a certain degree of improvement in decreasing the number or radical errors. A similar principle could also be used to explain another part of our results; that is, controlling for the extending characters’ structure could save students’ energy that could be allocated to other information about those characters, such as their multiple pronunciations. Thus, it is unsurprising to find that those who received visuospatially controlled CSDI had the fewest homophone errors.

Dictation and reading are two interconnected yet distinct skills in language learning (Karpinski, 2000), particularly in the context of Chinese character acquisition (Pan & Lin, 2023). The process of dictation—writing down words or characters spoken aloud—offers a unique lens into language acquisition, and its role in Chinese character reading is multifaceted (McBride-Chang et al., 2011). The process of dictation involves active listening, auditory processing, memory recall, and fine motor skills for writing (Cheng-Lai et al., 2013). Verbally controlled CSDI, which might involve the processing of breaking down complex characters into their simpler components in one’s mind, can significantly enhance the dictation process. This suggests that auditory and linguistic processing, combined with the cognitive skills of deconstructing and reconstructing Chinese characters, plays a crucial role in dictation.

This study has at least two limitations. First, the timing of the pretest, posttest, and training implementation was somewhat unpredictable, and the allotted time for data collection and training was limited. Most of the time slots selected were before the first class, after lunch (which is typically used for nap time and lasts approximately 30 to 40 min), and after school. These varying time slots could potentially have a biased effect on the training outcomes. Furthermore, several uncontrollable conditions could have influenced the instructions, such as students arriving late to school, engaging in activities in their classrooms (e.g., story time before their first class), or being too tired after school. The examiners were instructed to check whether the participants were in an appropriate state to participate in the training; in several cases, they recommended postponing the training due to the students’ conditions.

Furthermore, although we focused on examining the effectiveness of the different forms of CSDI on Chinese reading and writing in terms of the sounds, appearances, and amounts of extending characters, it is undeniable that relevant metalinguistic abilities, such as orthographic knowledge, phonological awareness, or morphological awareness, could also be influential, as we discussed earlier. However, due to the limited resources and spaces, the participants’ metalinguistic abilities were not involved in the analyses in the current study, which might lead to some uncontrolled impact. Thus, future studies are encouraged to consider this issue.

Despite these limitations, this study has again confirmed the unique importance of CSDI designs that could benefit Chinese character reading and dictation in Chinese children with dyslexia, and it is among the first to provide evidence concerning this issue. Furthermore, the aforementioned impacts were demonstrated primarily in that controlling for the extending characters’ pronunciations taught in each session could be more influential than controlling for their structure or not controlling, except for the benefits in decreasing homophone errors.

A salient feature of the study's outcomes is the revelation that the control of extending characters’ pronunciations during each instructional session manifests a more pronounced impact than controlling for their structural attributes. Interestingly, this control was beneficial across multiple measures, although it was particularly influential in reducing the frequency of homophone errors. Our findings could contribute to refining instructional techniques. In sum, the evidence furnished by this study not only adds to the cumulative evidence on the effective instructional designs for Chinese children with dyslexia but also delineates new directions for refining and tailoring educational strategies. Thus, the present investigation holds significant implications for educational psychologists, special education practitioners, and policymakers striving to provide optimal learning conditions for children with dyslexia within the Chinese linguistic context.

Supplemental Material

sj-docx-1-ecx-10.1177_00144029231220308 - Supplemental material for The Effectiveness of Diverse Designs of Chinese Stem-Deriving Instruction for Chinese Children With Dyslexia

Supplemental material, sj-docx-1-ecx-10.1177_00144029231220308 for The Effectiveness of Diverse Designs of Chinese Stem-Deriving Instruction for Chinese Children With Dyslexia by Li-Chih Wang, Duo Liu, He-Hsiu Lin, Kevin Kien-Hoa Chung and Zhengye Xu in Exceptional Children

Footnotes

Authors’ Note

Due to the nature of this research, participants of this study did not agree for their data to be shared publicly, so supporting data are not available. This study was supported by a grant from the Education Bureau of the Hong Kong Special Administrative Region, China (EDB/QEF/2018/0431), and the Ministry of Education, Taiwan (Yushan Young Fellow Program).

ORCID iDs

Li-Chih Wang

Zhengye Xu

Supplemental Material

Supplemental material for this article is available online.

References

Abbott

R. D.

Berninger

V. W.

Fayol

(2010). Longitudinal relationships of levels of language in writing and between writing and reading in grades 1 to 7. Journal of Educational Psychology, 102(2), 281–298. https://doi.org/10.1037/a0019318

Baddeley

(1986). Working memory. Clarendon Press/Oxford University Press.

Badian

N. A.

(2005). Does a visual-orthographic deficit contribute to reading disability? Annals of Dyslexia, 55(1), 28–52. https://doi.org/10.1007/s11881-005-0003-x

Barker

T. A.

Torgesen

J. K.

Wagner

R. K.

(1992). The role of orthographic processing skills on five different reading tasks. Reading Research Quarterly, 27(4), 334–345. https://doi.org/10.2307/747673

Brennan

McDonagh

Dempsey

McAvoy

(2022). Cosmic sounds: A game to support phonological awareness skills for children with Dyslexia. IEEE Transactions on Learning Technologies, 15(3), 301–310. https://doi.org/10.1109/TLT.2022.3170231

Chan

D. W.

C. S. H.

Tsang

S. M.

Lee

S. H.

Chung

K. K.

(2007). Prevalence, gender ratio and gender differences in reading-related cognitive abilities among Chinese children with dyslexia in Hong Kong. Educational Studies, 33(2), 249–265. https://doi.org/10.1080/03055690601068535

Chen

H. C.

Hsu

C. C.

Chang

L. Y.

Lin

Y. C.

Chang

K. E.

Sung

Y. T.

(2013). Using a radical-derived character e-learning platform to increase knowledge of Chinese characters. Language Learning & Technology, 17(1), 89–106. https://www.lltjournal.org/item/10125-24511/

Cheng-Lai

Li-Tsang

C. W.

Chan

A. H.

A. G.

(2013). Writing to dictation and handwriting performance among Chinese children with dyslexia: Relationships with orthographic knowledge and perceptual-motor skills. Research in Developmental Disabilities, 34(10), 3372–3383. https://doi.org/10.1016/j.ridd.2013.06.039

Chung

K. K.-H.

(2016). Understanding developmental dyslexia in Chinese: Linking research to practice. Asia Paciﬁc Journal of Developmental Differences, 4(1), 3–15. https://doi.org/10.3850/S2345734117000021

10.

Chung

K. K.

Mcbride-Chang

Wong

S. W.

Cheung

Penney

T. B.

C. S. H.

(2008). The role of visual and auditory temporal processing for Chinese children with developmental dyslexia. Annals of Dyslexia, 58, 15–35. https://doi.org/10.1007/s11881-008-0015-4

11.

Chung

K. K.-H.

Tong

McBride-Chang

(2012). Evidence for a deficit in orthographic structure processing in Chinese developmental dyslexia: An event-related potential study. Brain Research, 7(1472), 20–31. https://doi.org/10.1016/j.brainres.2012.06.010

12.

Ehri

L. C.

(1997). Learning to read and learning to spell are one and the same, almost. In Perfetti

Rieben

Fayol

(Eds.), Learning to spell: Research, theory, and practice across languages (Vol. 13, pp. 237–268). Lawrence Erlbaum.

13.

Elgün-Gündüz

Akcan

Bayyurt

(2012). Isolated form-focused instruction and integrated form-focused instruction in primary school English classrooms in Turkey. Language, Culture and Curriculum, 25(2), 157–171. https://doi.org/10.1080/07908318.2012.683008

14.

Flaugnacco

Lopez

Terribili

Montico

Zoia

Schön

(2015). Music training increases phonological awareness and reading skills in developmental dyslexia: A randomized control trial. PloS One, 10(9), e0138715. https://doi.org/10.1371/journal.pone.0138715

15.

C. S.-H.

Bryant

(1997). Learning to read Chinese beyond the logographic phase. Reading Research Quarterly, 32, 276–289. https://doi.org/10.1598/RRQ.32.3.3

16.

C. S.-H.

Chan

D. W. O.

Lee

S. H.

Tsang

S. M.

Luan

V. H.

(2004). Cognitive profiling and preliminary subtyping in Chinese developmental dyslexia. Cognition, 91(1), 43–75. https://doi.org/10.1016/s0010-0277(03)00163-x

17.

C. S.-H.

Chan

D. W. O.

Tsang

S. M.

Lee

S. H.

(2002). The cognitive profile and multiple-deficit hypothesis in Chinese developmental dyslexia. Developmental Psychology, 38(4), 543–553. https://doi.org/10.1037//0012-1649.38.4.543

18.

C. S.-H.

T.-T.

W.-K.

(2003). A “radical” approach to reading development in Chinese: The role of semantic radicals and phonetic radicals. Journal of Literacy Research, 35(3), 849–878. https://doi.org/10.1207/s15548430jlr3503_3

19.

C. S.-H.

Yau

P. W.-Y.

(2003). Development of orthographic knowledge and its relationship with reading and spelling among Chinese kindergarten and primary school children. In McBride-Chang

Chen

H.-C.

(Eds.), Reading development in Chinese children (pp. 51–71). Praeger.

20.

Y.-C.

(2001). A comparative study of the effectiveness of different instructional strategies of Chinese characters on third-grade elementary students with reading disabilities [text in Chinese]. Journal of Pingtung Teachers College, 14, 179–218. https://tpl.ncl.edu.tw/NclService/JournalContentDetail? SysId=A01017227&sysid=%2C&q%5B0%5D.f=JT&q%5B0%5D.i=%E5%B1%8F%E6%9D%B1%E5%B8%AB%E9%99%A2%E5%AD%B8%E5%A0%B1&q%5B1%5D.o=0&q%5B1%5D.f=*&q%5B1%5D.i=%E8%83%A1%E6%B0%B8%E5%B4%87&q%

21.

Y.-C.

(2005). The correlations among the related variables and the learned performances of four different Chinese characters teaching methods on fourth-grade elementary students with word recognition difficulties [text in Chinese]. Journal of Pingtung Teachers College, 22, 1–40. https://tpl.ncl.edu.tw/NclService/JournalContentDetail? SysId=A01017227&sysid=%2C&q%5B0%5D.f=JT&q%5B0%5D.i=%E5%B1%8F%E6%9D%B1%E5%B8%AB%E9%99%A2%E5%AD%B8%E5%A0%B1&q%5B1%5D.o=0&q%5B1%5D.f=*&q%5B1%5D.i=%E8%83%A1%E6%B0%B8%E5%B4%87&q%

22.

Huang

H. S.

(2001). Chinese character recognition test [text in Chinese]. Psychological Publishing.

23.

Huang

H. S.

Hanley

J. R.

(1995). Phonological awareness and visual skills in learning to read Chinese and English. Cognition, 54(1), 73–98. https://doi.org/10.1016/0010-0277(94)00641-W

24.

Hung

L.-Y.

Chang

Y.-W.

Chen

H.-F.

Y.-T.

Chen

C.-S.

(2003). The Battery of the Chinese of Pupils (基本讀寫字綜合測驗) [text in Chinese]. Psychological Publishing.

25.

Karpinski

G. S.

(2000). Aural skills acquisition: The development of listening, Reading, and performing skills in college-level musicians. Oxford University Press.

26.

Law

W. W.

Chung

A. L. S.

P. Y.

Lam

H. C.

(1998). Children’s stroke sequence errors in writing Chinese characters. Cognitive processing of the Chinese and the Japanese Languages, 113–138. https://doi.org/10.1023/A:1008091730338

27.

Law

S. P.

(2001). A case study of acquired dyslexia and dysgraphia in Cantonese: Evidence for nonsemantic pathways for reading and writing Chinese. Cognitive Neuropsychology, 18(8), 729–748. https://doi.org/10.1080/02643290143000024

28.

Lin

S.-J.

(1997). A comparison of the ability of first graders in Taiwan to learn graphemically-similar versus graphemically-dissimilar Chinese characters [text in Chinese]. Bulletin of Special Education and Rehabilitation, 6, 227–251. https://www.airitilibrary.com/Article/Detail/15633586-199806-201310010040-201310010040- 261-277

29.

Lin

Liu

(2019). Developmental changes in the relationship between character reading ability and orthographic awareness in Chinese. Frontiers in Psychology, 10, 2397. https://doi.org/10.3389/fpsyg.2019.02397

30.

M. C.

(2000). The effectiveness of CSDI on elementary students with severely word-recognition difficulties [text in Chinese]. Bulletin of Special Education, 18, 207–235. http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/6271

31.

Luo

Chen

Zhang

(2017). The inversion effect for Chinese characters is modulated by radical organization. Journal of Psycholinguistic Research, 46(3), 791–803. https://doi.org/10.1007/s10936-017-9484-0

32.

Luo

Wang

Zhu

Zhang

(2013). Working-memory training improves developmental dyslexia in Chinese children. Neural Regeneration Research, 8(5), 452–460. https://doi.org/10.3969/j.issn.1673-5374.2013.05.009

33.

Lyon

G. R.

Shaywitz

S. E.

Shaywitz

B. A.

(2003). A definition of dyslexia. Annals of Dyslexia, 53(1), 1–14. https://doi.org/10.1007/s11881-003-0001-9

34.

McBride-Chang

Chung

K. K.

Tong

(2011). Copying skills in relation to word reading and writing in Chinese children with and without dyslexia. Journal of Experimental Child Psychology, 110(3), 422–433. https://doi.org/10.1016/j.jecp.2011.04.014

35.

Meng

Hunnicutt

Seneff

Zue

(1996). Reversible letter-to-sound/sound-to-letter generation based on parsing word morpology. Speech Communication, 18(1), 47–63. https://doi.org/10.1016/0167-6393(95)00032-1

36.

Pan

D. J.

Lin

(2023). How do executive functions explain early Chinese reading and writing? Reading and Writing, 36(3), 625–-647. https://doi.org/10.1007/s11145-022-10314-1

37.

Peng

Namkung

Barnes

Sun

(2015). A meta-analysis of mathematics and working memory: Moderating effects of working memory domain, type of mathematics skill, and sample characteristics. Journal of Educational Psychology, 108(4), 455–473. https://doi.org/10.1037/edu0000079

38.

Piotrowski

Reason

(2000). The national literacy strategy and dyslexia: A comparison of teaching methods and materials. Support for Learning, 15(2), 51–57. https://doi.org/10.1037/10.1111/1467-9604.00145

39.

Rahimi

(2021). A comparative study of the impact of focused vs. comprehensive corrective feedback and revision on ESL learners’ writing accuracy and quality. Language Teaching Research, 25(5), 687–710. https://doi.org/10.1177/1362168819879182

40.

Raven

(1996). The Raven's progressive matrices: Change and stability over culture and time. Cognitive Psychology, 41(1), 1–48. https://doi.org/10.1006/cogp.1999.0735

41.

Rougier

P. R.

Bonnet

C. T.

(2016). How providing more or less time to solve a cognitive task interferes with upright stance control: A posturographic analysis on healthy young adults. Human Movement Science, 47, 106–115. https://doi.org/10.1016/j.humov.2016.03.001

42.

Schneider

Ennemoser

Roth

Küspert

(1999). Kindergarten prevention of dyslexia: Does training in phonological awareness work for everybody? Journal of Learning Disabilities, 32(5), 429–436. https://doi.org/10.1177/002221949903200508

43.

Shu

Chen

Anderson

R. C.

Xuan

(2003). Properties of school Chinese: Implications for learning to read. Child Development, 74(1), 27–47. https://doi.org/10.1111/1467-8624.00519

44.

Shu

Meng

Chen

Luan

Cao

(2005). The subtypes of developmental dyslexia in Chinese: Evidence from three cases. Dyslexia, 11(4), 311–329. https://doi.org/10.1002/dys.310

45.

Sun

W. C.

Yang

T. R.

Liang

C. C.

Hsu

P. Y.

Kung

Y. W.

(2006). The effects of computer assisted instruction to train people with reading disabilities recognizing Chinese characters. In Computer processing of oriental languages: Beyond the Orient. The research challenges ahead (pp. 139–148). Springer.

46.

Sweller

(1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. https://doi.org/10.1207/s15516709cog1202_4

47.

Tan

L. H.

Perfetti

C. A.

(1997). Visual Chinese character recognition: Does phonological information mediate access to meaning? Journal of Memory and Language, 37(1), 41–57. https://doi.org/10.1006/jmla.1997.2508

48.

Vitevitch

M. S.

(2008). What can graph theory tell us about word learning and lexical retrieval? Journal of Speech, Language, and Hearing Research, 51(2), 408–422. https://doi.org/10.1044/1092-4388(2008/030)

49.

Wan

Y.-Y.

(1991). Psychology for children’s Chinese character learning. In Yang

Z. F.

Gao

S. R.

(Eds.), Zhongguoren zhongguoxin: Fazhan yu jiaoxue pian (pp. 403–448). Yuan-Liou.

50.

Wang

C.-C.

(2005). Review of literature on character and vocabulary instruction studies in Taiwan [in Chinese]. Bulletin of Educational Psychology, 44(2), 253–272. https://doi.org/10.6251/BEP.20111109.

51.

Wang

L.-C.

Yang

H.-M.

(2014). Classifying Chinese children with dyslexia by dual-route and triangle models of Chinese reading. Research in Developmental Disabilities, 35(11), 2702–2713. https://doi.org/10.1016/j.ridd.2014.07.001

52.

Wang

L.-C.

Yang

H.-M.

(2018). Temporal processing development in Chinese primary school-aged children with dyslexia. Journal of Learning Disabilities, 51(3), 302–312. https://doi.org/10.1177/0022219416680798

53.

Wong

S.-H.

Sio

C. M.-F.

Leung

S.-O.

(2007). An exploratory study of “words by words” teaching method on primary two students’ competence and interest of learning Chinese words. Educational Research Journal, 22(1), 111–133. Retrieved January 22, 2023, from https://repository.um.edu.mo/handle/10692/35379

54.

W.-J.

(2007). Improving literacy teaching strategies and examples for students with reading disabilities in elementary schools [in Chinese]. Network Sociology Letters, 65. http://www.nhu.edu.tw/~society/e-j/65/65-43.htm

55.

Yang

Zhai

Zhang

Zhao

Weng

(2022). Prevalence of developmental dyslexia in primary school children: A systematic review and meta-analysis. Brain Sciences, 12(2), 240. https://doi.org/10.3390/brainsci12020240

56.

Yeh

S. L.

Chou

W. L.

(2017). Lexical processing of Chinese sub-character components: Semantic activation of phonetic radicals as revealed by the Stroop effect. Scientific Reports, 7(1), 1–12. https://doi.org/10.1038/s41598-017-15536-w

57.

Zhang

Wang

Wen

Qin

Zhong

(2006). A probabilistic feature based maximum entropy model for Chinese named entity recognition. In Computer processing of oriental languages: Beyond the Orient. The research challenges ahead (pp. 189–196). Springer.

58.

Ziegler

J. C.

Goswami

(2005). Reading acquisition, developmental dyslexia, and skilled reading across languages: A psycholinguistic grain size theory. Psychological Bulletin, 131(1), 3–29. https://doi.org/10.1037/0033-2909.131.1.3

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.82 MB