Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Influence of language immersion and phonological-based instruction on literacy skills of English language… Huo, Shuting 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_february_huo_shuting.pdf [ 3.33MB ]
Metadata
JSON: 24-1.0362561.json
JSON-LD: 24-1.0362561-ld.json
RDF/XML (Pretty): 24-1.0362561-rdf.xml
RDF/JSON: 24-1.0362561-rdf.json
Turtle: 24-1.0362561-turtle.txt
N-Triples: 24-1.0362561-rdf-ntriples.txt
Original Record: 24-1.0362561-source.json
Full Text
24-1.0362561-fulltext.txt
Citation
24-1.0362561.ris

Full Text

  INFLUENCE OF LANGUAGE IMMERSION AND PHONOLOGICAL-BASED INSTRUCTION ON LITERACY SKILLS OF ENGLISH LANGUAGE LEARNERS IN XI’AN, CHINA by SHUTING HUO B.Sc., Beijing Normal University, 2010 M.Ed., Beijing Normal University, 2013  A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Special Education)    THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)   December 2017  © Shuting Huo, 2017 ii  ABSTRACT Every child in China is required to learn English as a foreign language (EFL) in elementary school. They learn to not only speak the language but also read and write. Phonological awareness and phonological decoding are foundational skills to learning to read in English. Teaching these skills to young children can enhance their performance on word reading and spelling. Meanwhile, many Chinese children speak and understand very little English due to the lack of English language input in their daily lives. This research investigated the effectiveness of the phonological-based instruction, namely phonological awareness instruction and phonics, among Chinese children. I also examined the moderating effects of English oral language proficiency and language learning environment on the instructional effectiveness. Study 1 is a longitudinal study which followed 110 Chinese children who were learning English as a foreign language throughout an academic year in a school located in Xi’an, China, during which the students received systematic instruction on phonics. The participants were from two programs: one was an English immersion program in which intensive English instruction was delivered, and the other was the standardized program wherein students received less intensive English instruction. Developmental trajectories of two groups of students were compared. The findings indicated that the two groups improved at the same rate on real word reading, but students in the immersion program improved faster in pseudoword reading. Study 2 investigated the effectiveness of phonological awareness (PA) instruction in primary school EFL students. Eight classes from a school in Xi’an, China, were randomly assigned into two groups. Experimental classes received ten weeks of instruction, for a total of 380 minutes of instruction, on phonological awareness. The findings indicated that immediately after the instruction, the experimental group performed better at phonemic awareness and pseudoword reading. The  iii  instructional effects were not, however, sustained five months later after all the participants received phonics instruction. English expressive vocabulary at pretest significantly moderated the instructional effect on phonemic awareness. The results consistently suggested that Chinese children with better English oral proficiency and in enriched English language environment benefit more from the English phonological-based instruction.                 iv   LAY SUMMARY Every child in China is required to learn the English language in elementary school. One skillset that is fundamental to reading in English is the ability to analyze the speech into the smallest sound units, and map the sounds with letters. Teaching these skills to young children can help them become better readers. My research investigated whether this approach is effective among children in China. I first found that the skills noted above were very important in predicting English reading performance among Chinese children. Meanwhile, some Chinese children face the challenge that they speak and understand very little English because of the lack of English language input in their daily lives. I found that this had an influence on their gain from the instruction. Children with larger English vocabulary and in a better English language environment benefited more from the instruction.    v  PREFACE The author coordinated the teaching of the experimental program and data collection, and conducted the data analyses and the writing for this dissertation. Dr. Linda Siegel supervised the research program.  Ethics approval for this research project was granted by the University of British Columbia’s Behavioral Research and Ethics Board (UBC-BREB), certificate number H15-01369.   Portions of Chapter 2 and Chapter 5 have been published. Huo, S., & Wang, S. (2017). The effectiveness of Phonological-Based instruction in English as a foreign language in students at primary school level: A research synthesis. Frontiers in Education, 2(15), 1.  Partial results from Chapter 3 and Chapter 4 have been presented at a conference and published in the conference proceedings.  Huo, S., Siegel, L. (2016 May), Preparing Chinese children to learn to read English: Effects of phonological-based skill instruction In Y.Chi,. From class to the world-English immersion in China. Paper presented at the 6th China-Canada-the U.S. English Immersion Conference, Xi’an, China (45-80). Xian: Xian Jiaotong University Press Co.    vi  TABLE OF CONTENTS ABSTRACT .................................................................................................................................... ii LAY SUMMARY .......................................................................................................................... iv PREFACE ....................................................................................................................................... v TABLE OF CONTENTS ............................................................................................................... vi LIST OF TABLES ......................................................................................................................... ix LIST OF FIGURES ....................................................................................................................... xi ACKNOWLEDGEMENTS .......................................................................................................... xii CHAPTER 1: INTRODUCTION ................................................................................................... 1 English Immersion in China ............................................................................................... 2 Oral Language Proficiency and Phonological Awareness as Predictors of Word Reading 4 Effectiveness of Phonological-Based Instruction among ELL and EFL Students ............. 6 Learning to Read in English among Chinese-Speaking EFLs ............................................ 8 Purposes of This Research .................................................................................................. 9 Organization ...................................................................................................................... 10 CHAPTER 2: LITERATURE REVIEW ...................................................................................... 11 English Literacy Skills ...................................................................................................... 11 English Literacy Skills among Young EFL Students ....................................................... 16 Phonological Based Instruction ........................................................................................ 26 Literacy Education of Chinese and English in Lower Grades in China ........................... 33 Language Immersion ........................................................................................................ 36 Theoretical Framework ..................................................................................................... 41 Research Questions ........................................................................................................... 44 CHAPTER 3: STUDY 1 ............................................................................................................... 46 Introduction ....................................................................................................................... 46 Research Questions ........................................................................................................... 47 Method .............................................................................................................................. 48  vii  Results ............................................................................................................................... 57 Discussion ......................................................................................................................... 79 CHAPTER 4: STUDY2 ................................................................................................................ 89 Research Questions and Hypotheses ................................................................................ 89 Method .............................................................................................................................. 90 Results and Discussion ................................................................................................... 101 CHAPTER 5: GENERAL DISCUSSION AND CONCLUSIONS ........................................... 133 Development of English Phonological Awareness among Chinese EFL Children ........ 133 Influence of Phonological Awareness on English Reading among Chinese Speaking Children........................................................................................................................... 135 Effectiveness of Phonological-Based Instruction among EFL Students ........................ 137 Cross-Linguistic Transfer ............................................................................................... 140 Implications for Practice ................................................................................................. 142 Implications for Future Research .................................................................................... 143 Conclusions ..................................................................................................................... 145 BIBLIOGRAPHY ....................................................................................................................... 147 APPENDICES ............................................................................................................................ 158  Appendix A Sample lesson in phonics ........................................................................... 158 Appendix B English learning experience survey ............................................................ 160 Appendix C Chinese character identification ................................................................. 164 Appendix D Pinyin spelling items .................................................................................. 165 Appendix E Oral cloze items .......................................................................................... 166 Appendix F Picture naming items................................................................................... 167 Appendix G Word recognation tasks .............................................................................. 168 Appendix H Pseudoword reading items ......................................................................... 170 Appendix I Rhyme detection items................................................................................. 171 Appendix J Phonemic awareness tasks ........................................................................... 172 Appendix K Fixed effects of English real word reading ................................................ 173 Appendix L Fixed effects in model of pseudoword reading........................................... 174 Appendix M Fixed effects in model of simple pseudoword reading .............................. 175  viii  Appendix N Fixed effects in model of reading pseudowords with complex patterns .... 176 Appendix O Fixed effects in model of initial phoneme deletion .................................... 177 Appendix P Fixed effects in model of rhyme awareness ................................................ 178 Appendix Q Fixed effects in model of picture naming................................................... 179 Appendix R Fixed effects in model of oral cloze ........................................................... 180 Appendix S Fixed effects in model of Chinese character identification ........................ 181 Appendix T Partial correlation of all the outcome variables at three time points controlling for age in study 1 .......................................................................................... 182 Appendix U Chinese reading and spelling activities ...................................................... 184 Appendix V Vocabulary list of the experimental program ............................................. 185 Appendix W Sample lesson plan of PA instruction........................................................ 186 Appendix X Spelling scoring criteria ............................................................................. 187 Appendix Y Outline of the Pinyin program and a sample lesson ................................... 189 Appendix Z Correlation coefficients of the outcomes controlling for age in study 2  ... 190     ix  LIST OF TABLES Table 1 Means and standard deviations of home resources related to English learning ............................. 58 Table 2 Percentages and medians on the background variables of students from two programs ............... 58 Table 3 Means and standard deviations of all measures for students in both programs and both grades ... 61 Table 4 Concurrent correlation among English variables at time 1 and 2 .................................................. 67 Table 5 Correlation of predictors at time 1 and 2 with outcomes at time 2 and 3 among the immersion and non-immersion students .............................................................................................................................. 68 Table 6 Standardized fixed effects and significance in models of concurrent prediction of phonological decoding and oral language skills to word reading ..................................................................................... 69 Table 7 Standardized fixed effects and significance in models of longitudinal prediction of decoding skills to word reading ........................................................................................................................................... 70 Table 8 Standardized fixed effects and significance in models of longitudinal prediction of oral language skills to word reading .................................................................................................................................. 72 Table 9 Partial correlation between Chinese and English literacy outcomes controlling for age ............... 76 Table 10 Concurrent prediction of Chinese literacy outcomes to English literacy outcomes at time 1 and 2 .................................................................................................................................................................... 77 Table 11 The models of longitudinal prediction of Chinese outcomes to English literacy outcomes ........ 79 Table 12  Teachers’ background information ............................................................................................. 90 Table 13 Outline of the PA instruction ....................................................................................................... 95 Table 14 Demographic Characteristics of Participants before Program Implementation ......................... 101 Table 15 Means and standard deviations of the outcomes across teachers at three time points ............... 103 Table 16 Fixed effects in the model predicting word recognition. ........................................................... 107 Table 17 Fixed effects of pseudo word reading ........................................................................................ 109 Table 18 Fixed effects in model predicting initial phoneme deletion ....................................................... 111 Table 19  Fixed effect in the model of rhyme awareness ......................................................................... 113 Table 20 Fixed effects in the model of invented spelling ......................................................................... 114 Table 21 Fixed effects in the model of letter naming ............................................................................... 114 Table 22 Fixed effects in the model of oral cloze ..................................................................................... 115 Table 23 Fixed effects in the model of picture naming ............................................................................ 116 Table 24 Fixed effects in the model of Pinyin spelling ............................................................................ 116 Table 25 Fixed effects in mode of Chinese character reading .................................................................. 118 Table 26 Fixed effects in models predicting letter sound knowledge at delayed posttest ........................ 119 Table 27 Fixed effects in the models of spelling on the three elements ................................................... 125 Table 28 Partial correlation coefficients between predictors and outcomes after age was controlled ...... 127  x  Table 29 Fixed effects in models of phonemic awareness with Pinyin spelling, picture naming and word recognition as moderator ........................................................................................................................... 128 Table 30  Fixed effects in models of pseudoword reading with Pinyin spelling, picture naming and word recognition as moderator ........................................................................................................................... 129     xi  LIST OF FIGURES Figure 1 Estimated means of word reading for two programs at three time points .................................... 63 Figure 2 Estimated means of pseudoword reading for two programs at three time points ......................... 64 Figure 3  Cross-lagged correlation coefficients and standardized fixed effects between word reading and phonemic awareness ................................................................................................................................... 74 Figure 4 Cross-lagged correlation coefficients and standardized fixed effectsbetween word reading and rhyme awareness ......................................................................................................................................... 75 Figure 5 Data structure in study 2 ............................................................................................................. 101 Figure 6 Estimated means and confidence intervals of instructional outcomes at three time points controlling for English learning experience and teacher effect................................................................. 106 Figure 7 Instructional effect on pseudo word reading by teacher ............................................................. 110 Figure 8 Instructional effect on initial phoneme deletion by teacher ........................................................ 112 Figure 9 Pinyin spelling of students who fell behind the 50th percentile rank ......................................... 118 Figure 10 Average spelling response of each item of two groups at two time points .............................. 122 Figure 11 Percentage of correct responses on each element at posttest and delayed posttest .................. 123 Figure 12 Percentage of correct responses on each element of two groups at two time points ................ 124    xii  ACKNOWLEDGEMENTS First and foremost, I wish to thank the professors, principals, teachers and students involved in my work in Xi’an China. I wish to thank Professor Yanping Chi for putting me in contact with the school principals and her help in preparing the teaching materials. I thank the principals and English teachers in Shi-DA-Shi-Yan Elementary School and Lv-Di-Chan-Ba Elementary School for making it possible for us to work with the students. I also thank my research assistants from Xi’an Jiaotong University and Huaqing College for their diligent work in data collection. I wish to thank my supervisor, Linda Siegel, for giving me the excellent opportunity to study several interesting questions in bilingualism and language education as well as her great guidance throughout my PhD study. I would also like to thank my committee members Dr. Sterett Mercer and Dr. Lee Gunderson for their helpful comments. I want to thank Rick White from the Statistical Opportunity for Students (SOS) program for his invaluable suggestions regarding data analysis. I want to acknowledge that my PhD study was funded by the doctoral scholarship from The China Scholarship Council. I want to give special thanks to my boyfriend Sishuo Wang whose patience, love, and support made it possible for me to complete this journey. This dissertation is dedicated to my parents for their everlasting love and support. 1   CHAPTER 1: INTRODUCTION Learning the English language has been very popular and considered very important among Chinese people. Results of the most recent survey indicate that over 300 million Chinese people are learning or have learned English (Hu, 2005), making China the country with the largest population of English foreign language learners (EFL). English education in China is now universal and very important for academic success. The Chinese Ministry of Education has designated English language as a mandatory subject in primary school curriculum since 2001 (Ministry of Education China, 2001, 2011). The initial grade for English education is grade 3 generally, grade 1 in some areas where resources are available for providing the instructional environment and teacher training. English proficiencies are examined for enrollment of every level of education starting in secondary school. Test scores in the English language make up 20% of total score in the National College Entrance Examination, which is the most high-stakes test in China.   Recently English education in China has been under critical scrutiny. The teaching approach commonly adopted is grammar-translation method, focusing on language grammar and translating English language into Chinese. The method involves lots of paper and pencil activities and rote memorization. Other aspects of language such as conversational skills and listening comprehension are ignored. The dominant assessment method is using multiple choice questions rather than oral performance assessment. It has been considered as not being effective and inappropriate, especially for students at young age (Hu, 2005; Qiang & Siegel, 2012). Reform of English education has been advocated for many years. Actions have been taken to transform the education into communicative approach. The educational reform is reflected in the recently published national learning standards for English education (Ministry of  2  Education P.R. China, 2011). The goals of English instruction focus on communication. English education for young students should emphasize on cultivating students’ interest. According to the government endorsed textbooks, activities of practicing oral and listening skills increase in number, whereas paper pencil activities are not the most common for young learners. In addition, English reading instruction for students in early grades is not evidence based. No specific practices are suggested for the teachers. Whole word or “Look-and-say” approach, as opposed to a phonics based approach, is dominant in reading instruction in early grades (Yin, Anderson, & Zhu, 2007). The government curriculum does not encourage phonological awareness or phonics instruction. The concern is that the instruction is out of context and tedious. Thus children might lose interest in learning English at the very beginning. It is claimed in the document that explicit instruction on basic reading skills should not be introduced until children have developed a certain level of oral language proficiency (Ministry of Education P.R. China, 2011). Overall, the current public English Education at primary school level has undergone changes from the traditional approach to the communicative approach. At the same time, English immersion, the ultimate realization of the communicative approach, has been being experimented in some public schools in China in parallel to the dominant English language education. English Immersion in China  Language immersion, originated in Canada in 1960s, is a foreign language teaching approach. The approach aims to provide a communicative environment where the instruction emphasizes language use (Lightbown, Spada, Ranta & Rand, 2006). In immersion programs, instruction of subject content is mediated by the foreign language, so that students are given  3  authentic purposes to learn the language and that they are exposed to the maximized comprehensive input of the target language (Johnson & Swain, 1997). The subject content and classroom culture is parallel to the local curriculum and culture, so that students can draw from their skills and knowledge in their native language to learn the subject content and the new language. The skills and knowledge learned through the target language, in turn, can be applied in the local context and in the native language, thus the goal of additive bilingualism is achieved (Cummins, 1998). English immersion programs in China started in 1996, primarily for early grades from kindergarten to grade 3 (Qiang & Siegel, 2012). The idea and the implementation are transplanted from French immersion programs in Canada. Modifications are made due to curriculum constraints and limited resources. Hours of English immersion were cut down to 50% of the curriculum time (Qiang & Siegel, 2012). Teachers in the immersion program are Chinese native speakers. Despite all the limitations, English teaching in immersion classroom distinguishes itself from traditional English teaching non-immersion program. Instruction in immersion program is more intensive and the curriculum material is more diverse and advanced. Group and teacher-student interaction are greatly encouraged in class despite the limitation of large class size. In addition, instruction is fully mediated in English in the immersion program whereas Chinese is largely used as an aid in non-immersion program (Qiang & Siegel, 2012). The evidence indicates that English immersion program is an additive bilingualism program (Cheng, 2012). Students in the immersion program performed significantly better than students in non-immersion program on English oral language proficiencies, reading comprehension, word recognition and meta-linguistic awareness with no detriment to their  4  Chinese language skills and academic achievement (Cheng, Kirby, Qiang & Wade-Woolley, 2010; Knell et.al, 2007).  Typically, students in the immersion program in Xi’an received minimum instruction on phonics and no instruction on phonological awareness (PA). PA or phonics is not included in the current immersion program. Teachers have reported that students had problems in reading and understanding the content especially for students above grade 3, as the content becomes more difficult as it transitions from a global curriculum to subject-centered curricula which include social studies and sciences. Students need reading skills to acquire new vocabulary and understand the subject. In order for students to read to learn at the later stage, learning to read at the beginning stage becomes very important for the immersion program to be more effective.  Oral Language Proficiency and Phonological Awareness as Predictors of Word Reading Cognitive skills underlying word recognition are of two categories, language proficiency and phonological processing. Oral language proficiency is indicated by syntactic awareness and vocabulary (see Siegel, 1993 for a review). Vocabulary is the robust indicator of language proficiency and has a strong influence on word recognition and is therefore often controlled in studies on word recognition (e.g. Hulme et al., 2002). Vocabulary also has an indirect impact on word decoding by influencing phonological awareness. Children need to have a certain size of vocabulary to develop a linguistic insight of the phonological structure of the language (Metsala, 1999). A phonological task with unknown words measures working memory instead of phonological awareness (Siegel & Ryan, 1989). Syntactic awareness in part governs the process of predicting the meaning of an unfamiliar word in context. Evidence has shown that syntactic awareness, utilized by the tasks of oral cloze and oral correction, has unique influence on word recognition. This is one of the deficits of poor readers (e.g. Tunmer, Nesdale, & Wright, 1987),   5  and has predicted word reading among Grade 2 and 4 students after vocabulary and phonological memory were controlled (Cain, 2007).  Phonological processing skills are strongly related to word recognition at early stages of reading development. Phonological processing skills include phonological awareness and phonological decoding. Phonological decoding, which denotes in-depth analysis of the relation between the pronunciation and spelling of the word, is an essential skill in word recognition (Ehri, 1992; Perfetti & Hart, 2002). It is measured by the task of pseudoword reading to exclude the component of semantic retrieval.  Phonemic awareness is the most robust predictor of subsequent reading at early stage of learning to read. It refers to the ability of detecting and manipulating phonemes, the smallest linguistic sounds in speech, such as segmenting, deleting and blending (Hoien et al., 1995). Goswami (1988, 1991) argued that rhyme awareness is also an important factor in reading. Rhyme awareness underpins the strategy of analogy in word reading (Goswami, 1991). For example, “fought” can be analogous to the word "thought". To apply this strategy, awareness of rhyme units of the sound is necessary (Ziegler & Goswami, 2005). The recent meta-analysis of 235 studies found that phonemic awareness significantly predicted word reading more strongly than rhyme awareness and phonological memory (Melby-Lervåg, Lyster, & Hulme, 2012). The study also indicated that rhyme awareness independently contributed to word reading, although the pattern was less strong when compared to phonemic awareness. Phonological-based instruction, explicit instruction on phonological awareness and phonological decoding, has been found to have significant positive effects on word level reading.  The National Reading Panel (2000) recommended that explicit instruction of the phonological skills combined with letter knowledge should be included in literacy programs in early grades.  6  Meta-analysis of intervention studies of phonological awareness and phonics (Bus & IJzendoorn, 1999; Ehri, Nunes, Shanahan & Lesegremiums, 2001) found generally large to moderate instructional effects on word reading and spelling. Therefore, PA and understanding grapheme-phoneme correspondences are important fundamental skills. Effectiveness of Phonological-Based Instruction among ELL and EFL Students Studies have shown that phonological awareness and letter knowledge are important skills underlying word recognition among ELL (English language learners) and EFL students (e.g. Geva, Yaghoub-zadeh, & Schuster, 2000; Knell, 2007; Yeung & Chan, 2013). The National Reading Panel reported that ELL students generally respond to phonological-based instruction as well as the English native speaking students (August, Shanahan, & Escamilla, 2009). A synthetic review of studies with ELL students who are struggling readers (Richards-Tutor et al., 2015) showed that intervention programs effective on English native speaking children are also effective with ELL children regardless students’ language background.  Studies that investigated the effectiveness of the phonological-based instruction among non-native speakers were mostly conducted with ELL children who learn the English language in an English dominant society. Students who learn English as a foreign language are in completely different situations compared to the ELL students involved in the intervention studies in English speaking countries. EFL students are neither from language minority groups, nor are they struggling readers who fell behind English-native-speaking counterparts (Thorius & Sullivan, 2013). EFL students learn English as a school subject in their native countries. They have extensive print exposure of their first language and many have started to receive formal literacy instruction in their first language before they learn to read in English.   7  EFL students receive very little English exposure outside English class, thus the development of English oral proficiency is limited compared to ELL students in English dominant societies. The limited proficiency protracts development of word recognition among EFL children. Yin and colleagues (2007) found that grade 4 students in Beijing who had learned English for three years relied on visual cues to recognize words. The strategy of phonological decoding was not common until grade 6. In addition, limited language exposure and proficiency result in poor lexical representation, which means the form of words (spelling and pronunciation) is weakly connected to the meaning (Jiang, 2000). EFL students can learn to phonologically decode words, but have difficulty in retrieving the meaning because of limited vocabulary. Instructional experience of ELL and EFL students is also different in respect of receiving the phonological-based instruction. Firstly, most of the phonological-based instruction in the ELL studies were conducted in the remedial sessions to struggling readers, and conducted in a group or individual setting (Thorius & Sullivan, 2013). This rarely happens in the EFL context, because very few pull-out sessions could be dedicated to foreign language learning. Secondly, even there is evidence suggesting that phonological-based instruction is effective at whole-class level with ELL students (Stuart, 1999), the concept of whole-class setting can be very different across cultures. Teachers in the western context embedded small group and individual activities in whole class level instruction. In contrast, the whole-class level instruction in Asia is usually teacher-centered, and dominant with lecturing (Li, Rao, & Tse, 2012).  In addition, the role phonological-based instruction plays in ELL context may be different from the one in the EFL context. In the ELL context, phonological-based instruction is an important component in a comprehensive reading program for young children. It is supplementary to other instructional components including guided and shared reading, writing,  8  independent reading, and story-telling (D’Angiulli, Siegel, & Maggi, 2004). In the EFL context, phonological-based instruction is mainly adopted as a method to teach vocabulary (Lau & Rao, 2013; Lin & Cheng, 2008) and to correct pronunciation (Huang, Lin, & Su, 2004).   Considering the differences noted above, findings of the effectiveness of the phonological-based instruction with English native speaking and ELL children cannot be directly generalized to the EFL population. Evidence of effectiveness of the instruction with EFL children is needed.  Learning to Read in English among Chinese-Speaking EFLs Learning to read and spell in English is challenging for non-native children, especially for Chinese children who learn two drastically different orthographies.  The Chinese language is in the Sino-Tibetan language family, whereas English is in the Indo-European family (Shafer, 1955).  The Chinese language has phonological structure, which does not include consonant clusters; and the phonology-orthography correspondence happens only at syllable level (Siok & Fletcher, 2001). The Chinese language fosters development of syllable rather than phonemic awareness which is fostered by English (McBride-Chang, Bialystok, Chong & Li, 2004). Unlike other ELL students whose first language is similar to English, such as Spanish, Chinese students cannot draw from their first language experience to develop phonemic awareness (Yeong & Rickard Liow, 2012) which is essential for English reading. Therefore, explicit training of English phonemic awareness may be especially useful for this group of learners. In the meantime, Children in Taiwan and Mainland China, before learning to identify Chinese characters, learn a phonetic system (Pinyin in the Mainland and Zhu Yin Fu Hao in Taiwan) as a transition. Experience of learning the phonetic system in Chinese may influences their learning to read in English.   9  In summary, previous studies suggest the phonological based instruction is effective with English native and ELL students. EFL learners in mainland China have unique characteristics and learn English in various contexts, which might influence the extent of which students benefit from phonological-based instruction. Therefore, the instructional effectiveness in the EFL context in China needs to be validated. Purposes of This Research The current research has four purposes. First is to examine the relation between English phonological awareness and oral language abilities with English reading. Second is to examine the cross linguistic transfer of decoding skills by investigating prediction of English word reading from earlier Chinese literacy skills. The third purpose is to evaluate effectiveness of phonological-based instruction phonological awareness instruction and phonics instruction. The fourth purpose was to examine the influence of learning contexts (immersion vs non-immersion) on students’ learning of English reading, and to investigate its influence on concurrent and longitudinal predictions of English word reading.  To achieve the above purposes, two studies were conducted. Study 1 was a one-year longitudinal study comparing developmental trajectories of English reading skills of learning English in two contexts, immersion and non-immersion programs. All students from both programs received systematic instruction on synthetic phonics during the year. Study 2 was an intervention study investigating effectiveness of the phonological awareness instruction without using print in Chinese grade 1 students. Delayed instructional effectiveness was also examined after five months during which both groups of students received phonics instruction. The outcome measures were the same in two studies, including English phonological awareness, English word level reading and spelling, and English oral proficiency.  10  Association of prior Chinese literacy skills with later English reading abilities was examined in both studies. Study 1 investigated prediction of English reading one year later from Chinese literacy skills at time 1. The cross-linguistic transfer was compared across programs. Study 2 investigated the moderating effects of prior Chinese Pinyin spelling and character identification on the immediate instructional effects. Study 1 and study 2 both address the topic of cognitive-linguistic underpinnings of English word reading among Chinese EFL students. Study 1 focused on longitudinal and concurrent prediction of English word reading from English phonological processing skills, English oral proficiency and Chinese literacy skills. Study 2 adopted experimental manipulation to investigate the causal influence of phonological awareness on English word reading and spelling. It further investigated the moderating effect of English oral proficiency and Chinese literacy skills on this causal influence. Organization Chapter two presents literature review on the topic of research. Chapter three presents study 1 which investigated one-year development of English literacy skills of students from immersion and non-immersion programs. Chapter four presents study 2 which is an experimental study investigating effectiveness of phonological awareness training on English literacy skills. Chapter 5 presents discussion of the findings and implications for practice.    11  CHAPTER 2: LITERATURE REVIEW English Literacy Skills Word Recognition The act of reading is to derive meaning from print. Until fairly recently, written systems were considered to be a direct transcription of speech, with decoding considered the core of reading, and that learning to read was mastery of sound-print corresponding rules, such that the print form translates to pronunciation which then translates to meaning (See Gough & Tunmer, 1986 for a review). Goodman (1967) opposed this simple theory, and conceptualized reading as a psycholinguistic guessing game. Through analyzing errors children made in oral reading, he claimed that reading is an iterative process of predicting and self-correcting. Smith (1977) made a similar claim that good readers depend heavily on context cues to derive meaning of an unknown word, rather than relying on visual cues to recognize words. Thus the significance of word recognition in reading was undermined. The claim that precise letter and word recognition is not an essential process in reading has been disputed. Stanovich (1986) disputed the evidence that good readers rely more on prediction for comprehension rather than word recognition, stating that being able to make predictions and making use of context cues was the consequence of successful reading, rather than the cause. Independent and successful reading relies heavily on effortless and accurate word recognition. Through a substantial amount of research the conclusion has been reached that word recognition is the foundation of reading (Adams, 1990; Pefetti, 1995; Stanovich, 1986). Word recognition is defined as accessing a word’s meaning from its print form (Wolf & Katzir-Cohen, 2001). The seemingly easy process involves activation and interaction of multiple components, including phonological, orthographical and semantical processing (Perfetti & Hart,  12  2002; Seidenberg & McClelland, 1989). At least two components are involved in word recognition, word decoding and meaning access (Han, 2015; Perfetti & Hart, 1991). Word decoding refers to the process of retrieving phonological information from orthographic representation. Meaning access is to retrieve a words’ semantic information. To best capture the process of word reading, performance data on tasks that tap into both subcomponents should be adopted such as reading word out loud and lexical judgement (Harm & Seidenberg, 2004; Seidenberg & McClelland, 1989).  In the literature of development of reading abilities in children’s native language, word recognition and word decoding are often used interchangeably and measured by the same task of reading words out loud. Meaning access is not separately measured, because the test items employed are high-frequent and concrete words and supposed to be familiar to children (e.g. Ehri et al., 2001; Hoien, Lundberg, Stanovich, & Bjaalid, 1995). Tasks which measure semantic access such as lexical decision and word-picture matching are rarely used to measure word reading. Instead, semantic access is often measured to assess lexical knowledge or receptive vocabulary, and low frequency words are employed as test items (e.g. Nation & Snowling, 1998). Phonological Decoding Phonological decoding refers to the process where a reader derives meaning of a printed word by translating groups of letters to sounds and blending the sounds together to access the lexical representation in their oral vocabulary (See Ehri, 2014 for a review). One applies the skill of phonological analysis and knowledge of alphabetic principles to sound out a word instead of retrieving the information from memory using visual cues.  Phonological decoding is considered essential for word recognition in the dual-route model (See Coltheart, Rastle, Perry, Langdon, & Ziegler, 2001 for a review). The model  13  proposes two routes of retrieving meaning from print. Familiar and irregular words are recognized through the lexical route without phonological decoding. An unfamiliar word is identified through the non-lexical route by breaking it into phonological units which are then blended together to match the word from the reader’s mental lexicon. To account for the deep orthography of the English written system,  Ziegler & Goswami (2005) proposed that one has to master at least three strategies to identify most English words, which vary in size of the sound and print corresponding unit. The three strategies are decoding by grapheme-phoneme-correspondence rules, decoding by analogy using the shared rhyme constituent, and decoding by whole word recognition.  The idea that phonological decoding is necessary for word recognition is also advocated in the lexical quality hypothesis (Perfetti & Hart, 2002). According to the hypothesis, a word retrieved reliably and efficiently by sight is the one represented with good quality, meaning the word is represented with redundancy and specificity. Phonetic information of a word is redundant if the word’s pronunciation and spelling is already specifically represented. The redundant cues of phoneme-grapheme correspondence can confirm the connections among a word’s spelling, pronunciation and meaning by avoiding confusion with words similarly spelled or pronounced. For example, one who memorized the word “president” by rote and does not know how to phonologically decode, may have problems distinguishing it from words with similar visual forms such as “present”, “precedent”,” precede”. Spelling Spelling is strongly associated with reading. Ehri (1987) argued that the same set of cognitive abilities underlie both reading and spelling. Phonemic awareness and letter sound knowledge are also essential for spelling development (e.g. Hulme et al., 2012; Muter, Hulme,  14  Snowling, & Taylor, 1998). Spelling and reading development follow the same stages; the pre-alphabetic stage, partial decoding stage, decoding stage, and the consolidation stage. At the pre-alphabetic stage children have no awareness of the alphabet, their writing resembles drawings, or utilize random letter strings. At the partial decoding stage children can spell initial and final sounds accurately. At the decoding stage children can transcribe words phonetically, but are not able to follow the orthographic details such as digraphs and diphthongs. The final stage, consolidation, the child’s writing becomes refined. Children are able to spell words using larger represented units, such as rhymes and morphemes.  Before children reach the final stage and master English spelling, they produce errors that reflect their immaturity of phonemic awareness and insufficient knowledge of letter-sound conversions (Tangel & Blachman, 1992). Vowels are harder to spell than consonants, as a result of vowel sounds being aurally more similar to each other than consonants. Initial consonant spelling is developed earlier than final consonants (Lee & Al Otaiba, 2017; Tangel & Blachman, 1992). Spelling of letters in which the sound can be inferred from the letter names are more straight forward to spell than letters in which the name provides no suggestion (Bowman & Treiman, 2002). Children also make spelling errors in complex structures such as consonant clusters. These errors reflect their limited ability of phonological decoding (Treiman, 1991). Development of Phonological Awareness Grain size theory (Goswami & Ziegler, 2005) advocates that the linguistic complexity of English accounts for the development of phonological awareness.  Development of phonological awareness follows the descending order of grain size. Children first develop syllable awareness, then are able to detect the salience of the rhyme unit and develop the onset-rhyme awareness. Full development of phoneme awareness follows last, after children formally learn literacy.  15  Some evidence has shown that phonemic awareness is the consequence of learning alphabetic principles. Johnston, Anderson, and Holligan (1996) found that students acquire a basic level of alphabetic principles before they develop explicit phonemic awareness. Carroll (2004) found that students lacking any knowledge of letter sound cannot perform phonemic awareness tasks. Holm and Dodd (1996) found that adults who lacked the experience of learning alphabetic language had poorer phonemic awareness than their counterparts who have had the experience.  Besides linguistic complexity, performance of phonological awareness is also influenced by the type of task.  McBride-Chang (2004) argued that phonological awareness tasks which require explicit manipulation of speech sounds such as phoneme segmenting, blending, deletion and substitution, are harder to perform than tasks that rely on forced-choice judgement, including sound matching and “odd one out”.  In contrast to the grain size theory, which claims that phonemic awareness is the consequence of learning alphabetic principles, the lexical restructure model advocates that phonemic awareness is the precursor of learning alphabetic principles rather than the consequence (Metsala, 1999). A cross-linguistic study found that English and Czech speaking children were able to isolate phonemes in foreign languages, despite the fact that they had no prior knowledge of the letter transcription of the phonemes in the foreign language (Hulme, Caravolas, Málková, & Brigstocke, 2005). With rudimentary awareness of phonemes, children can infer letter sound from letter name. For example, the letter “T” is easier learned than the letter “W”, the name of which does not provide a source of inference to the sound the letter makes while utilized within words (Foy & Mann, 2006; Kim, Petscher, Foorman, & Zhou, 2010). A cross-lagged panel analysis was also conducted in the study to investigate the direction of the  16  relation between phonemic awareness and letter sound knowledge (Foy & Mann, 2006). The results indicated that the earlier phonemic awareness accessed by phoneme judgement was more closely associated with later letter sound knowledge than the reverse. This indicates a probable causal influence of phonemic awareness on the acquisition of letter sound knowledge.  The lexical restructuring model also advocates that the segmental representation of words emerges through growth from vocabulary rather than literacy experience. The development of phonemic awareness is preceded by implicit phonemic representation which is significantly influenced by the size of one’s spoken vocabulary (Walley, Metsala, Victoria, & Garlock, 2003). The Matthew Effect Reading achievement in later developmental stages is largely determined by early literacy skills (e.g. Cunningham & Stanovich, 1997). Students who were struggling with word recognition and spelling in their early years tend to fall further and further behind as they develop. This widening gap was termed the Matthew Effect in literacy achievement by Stanovich (1986). Young readers who master the skill of decoding the print-sound code tend to acquire more reading experience and are more likely to be exposed to words with complex spelling patterns. Through reading, children learn new words, acquire a larger size of vocabulary, learn syntax rules and gain general cognitive abilities. Skills and knowledge students learn though reading leads to further successes at reading at advanced levels. English Literacy Skills among Young EFL Students Foreign language learners vary drastically at the ultimate proficiency of the target language. Factors of two categories explain the individual differences, affective factors and aptitude factors.  Affective factors include variables such as motivation, willingness to communicate, anxiety experienced with the foreign language, etc. According to Lamb's (2011)  17  model of the Matthew effect in EFL learning, the home environment as well as the parents’ attitude influence students’ attitude towards English learning. Students with a positive attitude are more active in class and receive more attention from the teacher, thus have more experience with the language and develop better abilities, which in return reinforces the positive attitude. Although affective factors are important to foreign language learning, some researchers argue that they are not specific to language learning and are pervasive in learning other subjects (Ganschow, Sparks, & Javorsky, 1998). Aptitude factors that are specific to language learning significantly explain individual differences in foreign language achievement after affective factors have been partialled out. Researchers draw parallels between theories explaining individual differences in reading in the first language with learning a foreign language (Sparks, 1995, 2006). Word level skills, which are important for reading among English-native-speaking students, are also important among EFL students.  Word Recognition among EFL Learners The EFL learning environment is marked by limited English language input. The constrained language environment leads to delayed development of English word level skills for EFL students compared to English-native-speaking children. For example, the logographic stage of word reading is short-lived for native speakers, who develop beyond the initial stage soon after formal schooling starts (Ehri, 1987, 2005; Frith, 1985), but may last longer for EFL students. Yin et al. (2007) conducted a study in several schools in Beijing where English phonics was not systematically taught. They found that 50% of Grade 2 EFL students and 34% of Grade 4 EFL students in the sample were designated as recognizing words in pre-alphabetic stage, which indicates that they recognized print words using visual features instead of phonological decoding.  18  The EFL context not only has an impact on the development of English word recognition, it also influences its structure. English word recognition is a multi-facet concept in EFL learners. The two subcomponents entailed in word reading, meaning access and word decoding, are shown to be dissociated in EFL learners, especially among low proficiency learners. Jiang (2000) proposed that formal representation (spelling and pronunciation) of a word is weakly linked to semantic representation of the word among low proficiency EFL learners, because English words are not acquired naturally for this group of learners.  Foreign language teachers introduce new words by directly pairing the pronunciation and spelling with pictures or L1 translations and ask students to repeat and spell the words several times (e.g., De Groot & Keijzer, 2000). In this process, emphasize is put on learning a word’s pronunciation and spelling; students are provided with limited opportunities to acquire the word’s meaning naturally through multiple exposure. The dissociation of word decoding and semantic access among low proficiency EFL learners was evident in the following studies. Shiotsu (2009) measured word decoding by visual form matching and meaning access by synonym/antonym decision. The results indicated that the skilled readers showed smaller discrepancies between the two tasks than lower-level readers. Saiz (2007) fitted the data of word recognition in Moroccan children who just began to learn Spanish in multivariate models. The results showed that word decoding (indicated by real word pronunciation and pseudoword reading) and semantic access (indicated by picture matching and first language translation), were separate psychological constructs.  The association between word decoding and semantic access is influenced by learners’ first language background. Learners with logographic first language background shower greater dissociation between the two aspects of word recognition. Brown and Haynes (1985) measured oral pronunciation and visual discrimination of English words among Japanese and Arabic native  19  speakers. Japanese-English speakers performed the best in visual discrimination of English words but more poorly in oral pronunciation tasks. Hamada and Koda (2008) adopted the paired-learning paradigm to Korean and Chinese L1 learners, and found that semantic acquisition of words is more associated with phonological decoding for Korean learners but not for Chinese learners.  Spelling among EFLs Studies of English language learners show that their spelling errors generally follow the same pattern as that of English native speaking children. Simultaneously constrains of first languages are also revealed (see Figueredo, 2006 for a review) . ELL students, in the initial phase of learning to spell in English, present with difficulty with the phonemes that are absent in their native language. For example, Cantonese speaking children have difficulty spelling the phoneme “th” which does not exist in the Cantonese phoneme inventory, when compared to spelling other digraph consonants (Wang & Geva, 2003). EFL children also have complications spelling the structure that contradicts with their native language. For example, Taiwanese children spell final consonants poorly because they have no experience with final consonant stops in Mandarin speech (Hong & Chen, 2011). Influence of Phonological Awareness on Literacy Skills Some researchers argue that word-level reading is script dependent, meaning it largely depends on knowledge and orthographic skills of the target language. English phonological awareness explained a considerable amount of variance of English word identification after background variables, Chinese literacy skills, and English oral proficiency were statistically controlled both in an ELL context (Geva et al., 2000) and an EFL context (Chow, McBride-Chang, & Burgess, 2005; Knell et al., 2007).   20  Other than phonological awareness, letter sound knowledge is also an important underlying skill of English reading for Chinese children who learn English as a foreign language.  Letter sound knowledge uniquely contributes in explaining spelling performance of Mandarin L1 children in Singapore after the influence of phonemic awareness has been controlled (Yeong & Rickard Liow, 2011). The same influence was found within children in Xi’an, China, where a study showed that letter naming predicted word reading independent of phonemic awareness (Knell et al., 2007). There are a small number of experimental studies which investigated the causal influence of the phonological-based variables on English reading and spelling. The studies are reviewed and synthesized later in this chapter. Influence of Oral Proficiency on Literacy Skills The National Reading Panel (Geva, 2006) reviewed the empirical researches on second language proficiency and second language reading among ELL students within North America , and found that oral proficiency predicted less variance of word level reading (3% to 4%) compared to phonological processing skills. The research indicated weak association between oral language proficiency and word level skills. ELL students in early grades were equivalent to the English native speaking students at word level reading and spelling, despite that they lagged behind in oral language proficiency. The disadvantage of lack of language input is reflected in reading in higher grades, such that ELL students with smaller vocabulary tend to fall behind in reading comprehension. Oral language proficiency does not have an impact on word recognition among ELL students in Canada (Geva et al., 2000). This may be due to the success of literacy education in the early grades. Regardless of their language status, students received intensive and comprehensive literacy programs, where phonological-based instruction is a significant  21  component within these literacy programs. ELL students can learn to decode words very well despite the disadvantage of lack of language input.  Despite studies with ELL learners found that oral proficiency explains limited amount of individual differences of English word recognition at the beginning stage, studies with EFL learners, who generally learn English in a language-limited environment and have lower language proficiency, show that English oral proficiency plays an important role in word recognition among young learners. Kahn-Horwitz, Shimron, and Sparks (2006) compared good English word readers and the poor readers among children in Israel. The results showed that the two groups differed significantly on English vocabulary. A study with EFL learners in Beijing found oral vocabulary is important to all measures of word recognition after phonological processing, first language abilities and general intelligence have been controlled (Liu & Tao, 2007).   Cross-linguistic Transfer of Word Decoding The linguistic coding difference hypothesis (LCDH) advocates that the same set of cognitive skills which underlie the reading process in one’s first language also predict reading performance in a foreign language (Ganschow et al., 1998; Kahn-Horwitz, Shimron, & Sparks, 2005). These skills include phonological awareness, synthetic awareness, and morphological awareness. Hu and Schuele (2015) found that students who were experiencing difficulty learning a foreign language had mild deficits in meta-linguistic skills in their first language. Longitudinal studies suggest that prior literacy skills and print exposure in L1 predicted later literacy performance in a foreign language (Kahn-Horwitz et al., 2005; Shum, Ho, Siegel, & Au, 2016). According to the LCDH, students utilize their knowledge and skills in L1 in learning to read in a new language. Early attainment of reading in a second language largely depends on  22  students’ skills and knowledge in their first language. Studies with immigrant students in Canada indicated that students who have a history of learning literacy in their home country performed significantly better than students who had no such experience (Gunderson & Siegel, 2001).  The tenet of the LCDH is also embodied in the linguistic interdependent hypothesis (Cummins, 1979). It proposed that the meta-linguistic skills are language independent, and once established are transferrable across languages. Many studies found correlations between literacy abilities of ones’ first language and the second/foreign language. Melby-Lervåg and Lervåg (2011) conducted a meta-analysis of studies on the topic of cross-linguistic transfer among bilingual students at the primary school level, which found a moderate and reliable cross-linguistic correlation between word decoding (r = 0.42) and phonological awareness (r = 0.54). A correlation between L1 and L2 decoding skills was moderated by languages spoken as well as the language of instruction. It was found that this cross-linguistic correlation was stronger in samples where both L1 and L2 were alphabetic, than in samples where L2 was alphabetic and L1 was ideographic. The correlation was higher in samples where instruction was delivered in both children’s first and second language, than in samples where instruction was only delivered in one language.   Chinese languages have very different orthography and phonology from English, as learning to read and spell in English is particularly challenging for Chinese children.  Chinese written languages are logographic. Chinese speech is transcribed at the syllable level by characters which is a combination of strokes. Learning to read in Chinese at an early stage involves rote memorization and visual orthographic processing. On the other hand, correspondence between syllables and Chinese characters is not entirely arbitrary as Chinese characters do provide some phonological information. 78% of  23  Chinese characters are compound characters and have a phonetic component from which pronunciation of a character can be inferred (Shu, Chen, Anderson, Wu, & Xuan, 2003). For example, Chinese characters 其,期,棋,淇 are homophones (/chi2/) and they share the same phonetic feature 其 /chi2/. The pronunciation of the four characters is the same as the phonetic feature 其. In addition, occasionally the phonetic feature indicates a sub-syllabic unit. For example, the pronunciation of the Chinese characters岂/chi/, 己/ji/, 圮/si/ share the same rhyme unit /i/. Chinese languages and the English language have very different writing systems, thus it is expected that abilities of decoding Chinese characters will not correlate with the ability of decoding English words. There is evidence that supports the hypothesis indicating low and non-significant correlation among Chinese ELL students in Canada (Bialystok, McBride-Chang, & Luk, 2005; Gottardo, Yan, Siegel, & Wade-Woolley, 2001).  Surprisingly,  in contrast to Chinese-English bilinguals in English dominant societies like North America and Australia, a cross-cultural study found that the decoding skills of the two languages correlated moderately strongly among English language learners in Hong Kong (Bialystok et al., 2005). They further argued that the different patterns of cross-linguistic association were due to the difference in instructional experience and reading strategy. They believed that Chinese children in Hong Kong adopted visual-orthographic strategy rather than phonological decoding to read both Chinese and English script, as literacy instruction in Hong Kong adopted the whole word approach, which focused on visual word recognition, and did not encourage phonological disestablishment of words (Bialystok et al., 2005). The hypothesis was validated by the strong cross-linguistic correlation of word recognition found among low-proficiency bilingual children in Canada.  The correlation weakened in children with high proficiency, as they have learned to adapt decoding  24  strategies in accordance to script, phonologically in English and visually in Chinese (Bialystok et al., 2005).  However, studies have shown that phonological recoding also underlies reading in Chinese, and it plays a role no less important than visual-orthographic processing. So and Siegel (1997) found that tone discrimination and rhyming were the most important underpinning abilities of Chinese word recognition. Hu and Catts (1998) found that Chinese character reading was more associated with phonological awareness than visual memory for children in Taiwan.   The correlation between Chinese and English decoding skills could be contributed to the underlying phonological processing commonly shared in reading across languages other than visual memory. Hu (2013) found that, in her sample, Taiwanese students’ performances in Chinese reading and English reading correlated with each other in both grade 3 and grade 5. This indicated that the correlation was developmentally stable, and did not decrease as students’ English proficiency improved. Chinese character identification was also found to be associated with English word reading in Yin, Anderson, and Zhu (2007)’s study of students in Beijing. The association was significant even after the visual-inspection ability was controlled.  Role of Pinyin on Chinese-English Bi-Literacy  Pinyin is a phonological encoding system of Chinese languages. The encoding system uses Latin alphabets to represent Chinese speech. The Pinyin system resembles the English writing system such that they both feature grapheme-and-phoneme correspondence. Pinyin is a more transparent system and has symbols for tones. Another difference between the two systems is that letter sound in Pinyin is in syllables whereas in English the letter sound is in phonemes.  25  For example, the letter b is pronounced /bo/ in Pinyin, while it is pronounced /b/ in English phonics.  In mainland China, teaching of Pinyin is universal in primary school and usually taught right upon school entry. Students have mastered the skill by the end of the first or second year of school (Cheung & Ng, 2003).  Children who learn English in early grades in Mainland China will most likely be taught the two alphabetic systems simultaneously (Cheung & Ng, 2003; Ruan & Leung, 2012). The instruction of Pinyin enhances Chinese early literacy skills, such as phonological awareness, and character recognition. Shu, Peng and McBride-Chang (2008) found that Children's Chinese phonological awareness increased dramatically (74%) from kindergarten to grade 1 after the Pinyin instruction was delivered. Children from Taiwan and Mainland China, where the phonetic system is taught, outperformed children from Hong Kong on Chinese onset-rhyme awareness (Cheung, Chen & Yip, 2001).   The application of skills required for the Chinese phonetic systems can enhance English phonemic awareness, as Indicated by several studies comparing Chinese children with and without experience of learning Pinyin on their English phonemic awareness. Before they learn to read, children from the Mainland performed equally well with children from Hong Kong at all levels of phonological awareness. After the formal literacy instruction started, grade 1 students from the Mainland performed better than children from Hong Kong on syllable awareness and phonemic awareness (Cheung, Chen, & Yip, 2001; McBride-Chang et al., 2004). The advantage of learning the phonetic system on the development of English phonological awareness, continued to the secondary school level (Mcdowell & Lorch, 2015) and subsequent adulthood (Holm & Dodd, 1996).   26  The Pinyin skill is also shown to be transferable to English word reading and spelling skills, as Pinyin naming was correlated with English pseudo word reading, phoneme deletion and regular word spelling (Wang, Perfetti & Liu, 2005). The results of a cross-cultural study showed that phonemic awareness was more related to English word reading among children from Mainland China, whereas syllable awareness was the most powerful predictor of English word reading for children from Hong Kong (McBride-Chang et al., 2004). Phonological Based Instruction  Definition of Phonological-Based Instruction Phonological-based instruction is an approach of teaching children to read. It focuses on direct and explicit teaching of phonological analysis of words and converting the sound units to print. Phonological-based instruction includes two types of instructional programs, phonics instruction and phonemic awareness (PA) instruction. The phonological based approach is a contrast to the whole word approach, which instructs children to recognize a word as a whole without breaking it down to its letters or letter groups. The phonological based approach also contrasts whole language instruction, in which reading is taught by authentic reading and word recognition and phonological skills are not explicitly taught. Phonics instruction focuses on explicit and direct teaching of alphabetic principles, including letter naming, letter sound knowledge, and word decoding, which is to break a word down into letters and map them to corresponding phonemes to sound out the word. Applying this skillset to word recognition and text-level reading is also included in the instruction.  PA training focuses on phonemic awareness. Games and other activities are employed that access phonemic awareness, such as identifying, categorizing, segmenting, and blending phonemes. Other outcomes targeted by this instruction include rhyme awareness and syllable  27  awareness as the prerequisites to phonemic awareness. Implicit rhyme awareness is fostered through activities like singing and chanting nursery rhymes and poems. The purpose of such training is to direct students' attention from the meaning of words to the sounds of words.  Another purpose to include rhyme awareness as a learning outcome is that rhyme awareness underpins the ability of using analogy in word reading and spelling (Goswami, 1991). Syllables are larger than phonemes, and as such children develop syllable awareness prior to phonemic awareness. Children practice deleting, counting, segmenting, and blending syllables, the sound units that they are already aware of, to familiarize themselves with the operations which further can be applied on phonemes. There is overlap between phonemic awareness training and phonics instruction (Ehri, Nunes, Willows, et al., 2001). Both forms of instruction may include the component of grapheme-phoneme correspondence of 26 alphabetic letters. Phonics instruction goes beyond teaching letter sound knowledge to more complex content such as complicated spelling rules (digraphs and diphthongs) and word decoding. PA instruction focuses on detecting and manipulating speech sounds with little emphasis on written words, whereas phonics instruction focuses on deciphering the printed word. Word reading and spelling are important outcomes of phonics instruction, but not of PA instruction (Ehri, Nunes, Stahl, & Willows, 2001).  Alternative Approaches of Literacy Instruction Whole word approach. Advocates of the whole word approach believe that sight words are acquired through repetitive encounters rather than phonological decoding. The whole word approach manifests in different forms of deliveries, such as the “look-say” method, which utilizes flash cards to enhance rapid recognition and mnemonics to facilitate memorization of whole words. The whole word approach can also be delivered implicitly. Students are required to  28  read texts which contain controlled vocabularies and feature repetition of words such as the basal reader program (Gunderson, 2014).  A tenet in the whole word approach is that a basic level of sight word vocabulary is the prerequisite of phonics. Instruction of alphabetic principles should wait until students have a certain size of sight vocabulary, which is acquired through whole word recognition. Teaching spelling rules should be inductive. Sight words provide material for students to analyze and induce spelling rules. Whole language. Advocates of the whole-language approach believed that reading instruction should be meaning-and-strategy centered, and opposed the practice that phonics is delivered systematically and explicitly to children who are beginning readers (Goodman, 1986; Krashen, 2002; Smith, 1999). Empirical evidence of the effectiveness of the whole-language approach is not as much as that of the phonological-based approach. Gunderson (1997) believed the reason is that whole language is a propositional intertext rather than a specific teaching approach. The definition of whole language approach varies across contexts and even individuals. It manifests itself as intertext among a family of teaching approaches, including emergent literacy approach, meaning centered approach, and language-experience approach. The commonalities shared across the approaches are, 1) that speaking, listening, reading and spelling should be taught interactively and simultaneously; 2) that reading instruction should be directed to comprehension; 3) that learning of content and learning of language are inseparable. Whole language acknowledges the roles phonological skills and decoding play in reading, but direct teaching of letter sounds and phonemic awareness and skill-drill practices are not encouraged. Teachers should create opportunities for children to explore spelling rules on their own such as  29  invented spelling (Bergeron, 1990). The instruction on phonological skills must be embedded in meaningful, game-format and purposeful activities.  Outcomes Findings from quantitative meta-analysis studies indicated that phonological-based instruction has a moderate to significant effect on English literacy skills (Ehri, Nunes, Stahl, et al., 2001; Ehri, Nunes, Willows, et al., 2001). The overall effect size of phonemic awareness instruction is 0.86 on phonemic awareness compared to the conditions which did not contain a phonological component, 0.56 on pseudoword reading, 0.61 on word reading, 0.57 spelling and 0.34 on comprehension. The overall effect size of phonics instruction is 0.6 on pseudoword reading, 0.4 on word reading, 0.35 on spelling and 0.27 on comprehension.  The instructional effect is the same between English native speaking children and children whose first language is not English (D’Angiulli, Siegel, & Maggi, 2004; Geva, 2006; Lipka & Siegel, 2010).  Components of Phonological-Based Instruction Phonemic awareness, alphabetic principle, and phonological encoding and decoding of words are three major components featured in phonological-based instruction. In practice, phonological-based instruction varies greatly on components delivered and the sequence of delivery. Some teachers and schools deliver all the components simultaneously by incorporating them within one program. Some choose to introduce the components in a linear fashion. For example, phonemic awareness is delivered first as a prerequisite, followed by alphabetic principles and then word decoding. Some choose to remove certain components and allocate the time and resources to teaching other components. For example, phonemic awareness instruction might be removed for the sake of efficiency.    30  The National Reading Panel (2000) synthesized instructional conditions featuring different combinations of components, and compared effectiveness of different combinations. The results showed that the phonics program overall was more effective on word reading (d = 0.67) than phonemic awareness training (d = 0.53), suggesting instruction with phonological decoding and alphabetic principal is more effective on word reading than instruction which merely focuses on phonological skills. Phonological awareness training without manipulating letters is less effective than instruction combined with letters and print on reading and spelling, although inclusion of letters does not influence the instructional effectiveness on phonemic awareness (Ehri, Nunes, Stahl, et al., 2001). Castiglioni-Spalten and Ehri (2003) investigated the effectiveness of activities that were purely auditory without any use of print. The results showed that articulating phonemes in speech without using print had a positive but small effect on word reading for grade 1 struggling readers. Furthermore, training on implicit detection of phonemes was shown to have no effect on word reading. Instruction on alphabetic principles without explicit training on manipulating speech sounds shows a positive effect on phonemic awareness. Joseph (1999) found that word sort, a phonics activity which did not explicitly address phonological skills such as deleting and blending phonemes, significantly improved students’ phonemic awareness, in comparison to the control group. This suggests that students gained insights into the phonemic structure of the English language through learning letter sounds and phonological analysis of words, despite the fact that phonemic awareness was not directly taught.  Word decoding and encoding is another important component, as instruction on phonemic awareness instruction and letter sound knowledge is not sufficient for students to learn  31  to read words. Fielding-Barnsley (1997) found that students who are proficient in letter sound correspondence and phonemic awareness cannot automatically apply the grapheme-phoneme-corresponding rules in reading and spelling words. Explicit instruction on phonological encoding and decoding is still needed and would be beneficial. Whether PA activities can enhance the effects of phonics instruction has not been conclusively proven. There is evidence indicating phonics instruction is more effective with PA activities. Foorman et al. (2003) compared effectiveness of two types of phonics programs at the classroom setting, one program group contained few PA activities and the other were programs with many PA activities. The results indicate that phonics programs with high level of PA activities were more effective than phonics with no or few PA activities on literacy outcomes. Contradicting Foorman et all (2003),  Hatcher et al. (2006) found that that explicit phonemic awareness instruction was not necessary if instruction on phonological encoding and decoding was introduced. However, for young children at risk of reading delay, additional training in phoneme awareness and linking phonemes with letters was beneficial. Given all the research evidence that phonemic awareness is correlated with knowledge of alphabetic principles as well as early reading and spelling, it is natural to assume that phonemic awareness instruction should serve as a prerequisite to phonics instruction. Children who received PA instruction in kindergarten should be more prepared to learn alphabetic principles, reading, and spelling (Groff, 2001). However, there is little evidence that suggests influence of earlier phonemic awareness instruction on subsequent mastery of letter sound knowledge. No experimental study exits which compares two groups of learning alphabetic principles, of which one research group received PA instruction as a prerequisite, and the other has not.   32  Effectiveness of Phonological Based Instruction in the EFL context The concept of introducing phonological-based instruction in EFL classes is attracting ever more attention of researchers, school administrators, and teachers. Some believe that as EFL students do not have a natural language environment of English to develop the literacy skills spontaneously, the foundational skills such as phonemic awareness and alphabetic principles need to be explicitly taught so that students are prepared to learn to read in English (Shen, 2003). Government endorsed English curriculums in many EFL countries include phonics or PA as an instructional component, such as Malaysia (Johnson & Tweedie, 2010) and Taiwan (Lai, Tsai, & Yu, 2009). Phonological based instruction, in particular phonics instruction, has become a trend in English classes. Qualitative studies have found that some English teachers in EFL countries maintain the belief that these spelling rules (phonics) are essential in teaching English to young children (Kuo, 2011). For example, some English teachers in Hong Kong reported that they found phonics instruction effective in improving their students’ spelling and reading performances in early grades (Lau & Rao, 2013).    Compared to research studies conducted in English speaking countries, phonological based training studies are small in number in the EFL context. Huo (2017) systematically reviewed 15 experimental and quasi-experimental studies published between 2000 and 2016 on the topic of the effectiveness of phonological-based instruction in the EFL context. The results of this systematic review showed a consistent pattern that phonological-based instruction has positive effects on phonological decoding and phonemic awareness. However, the effectiveness found should be interpreted with great caution, because the casual inference in respect to the instructional effectiveness can be drawn in only three studies ( Dixon et al., 2011; Lai et al., 2009; Yeung et al., 2013). The rest of the studies are potentially biased for the reasons including failure  33  to exclude confound factors, mismatch between unit of assignment and unit of analysis, contamination and variation of the comparison conditions, inconsistency of outcome measures, and absence of required descriptive information. Furthermore, none of the three studies which qualified for making causal inference assessed reading with full validity; the semantic access of word recognition was not measured, although they showed positive effects on word decoding. Therefore, this study provides limited evidence for effectiveness of the phonological-based instruction on reading in English among young EFL children. Literacy Education of Chinese and English in Lower Grades in China  Chinese Literacy Education Through analyzing the national curriculum standards and teachers’ lesson plans, Wu, Li, and Anderson (1999) concluded that public literacy education in China is under great influence of national curriculum and textbook series, thus it is very standardized. Most children in China receive the same lessons in the same pace throughout the country. Primary schools in China are departmentalized, where usually one Chinese literacy teacher is responsible for one or two classes, containing 60 to 90 students. Students in lower grades usually receive 4-5 hours of Chinese literacy instruction per week and have 10-20 minutes of homework daily.  Literacy instruction in lower grades is skill-based. It is recommended that instruction on character recognition should precede reading instruction, as children should be taught to recognize around 2000 characters before being taught to read (Zhang, 1993). According to the national curriculum standards, students must be able to recognize 1600-2000 and write 800-1000 Chinese characters by the end of grade 2 (Ministry of Education P.R. China, 2011). Direct instruction, skill-drill practice, and pen-and-paper activities dominate Chinese lessons (Wu et al., 1999).   34  Reading comprehension is taught through teacher-led intensive reading, and the materials are mainly articles from the textbook series. They are composed of excerpts from classical literature and short essays, of which students learn 20-30 articles per term. During the instruction, every paragraph and sentence is analyzed in detail by teachers.  Students are then required to read the text aloud several times and answer comprehensive questions, and are required to recite the text occasionally. Writing and composing in response to the text is not required until grade 2. Extensive reading is downplayed in the curricula, and therefore independent reading of books of students’ choosing is not encouraged at school (Li, Rao, & Tse, 2012; Wu et al., 1999).  English Literacy Education Similar to Chinese literacy education, English literacy education in China is also under great influence of the national syllabus and text book series, as English is a mandatory subject in primary school curriculum. At the same time, there exists great disparities between rural and urban areas, as well as between coastal and interior areas (Nunan, 2003). English education starts in grade 1 in urban and coastal areas, and in grade 3 in less developed areas. English teachers’ English proficiency and education levels vary greatly.  Language and literacy education has undergone great reforms, through four versions of the national syllabus, of which the first was published after the foundation of PRC, and released in 1978, to the most updated version published in 2011. Contrasting with the oldest versions, authentic task based, self-regulated learning has been promoted in the new standard. Specific standards regarding reading have been added in the latest version. Another important distinction was that instruction of reading was included in the curriculum for learners in lower grades, whereas the older versions indicated that specific goals for reading were not mentioned until grade 6. The published goal of reading for the lower level (grades 3–4) instruction is as follows,  35  “Be able to recognize words printed with pictures, recognize and read aloud learned words with the ability to match words with their referents, and read and understand short picture stories. “  One important distinction is that word recognition is acknowledged as the foundation of reading in the updated version, which was treated as a component in vocabulary learning in the older versions.  Although the curriculum goals and content has evolved over the four editions, there have been challenges in implementation, such as a lack of print exposure and literacy support in schools, or a lack of experienced teachers and teacher training programs (Zhang, 2012). It was observed that the teaching method of English literacy is very similar to that of teaching Chinese. Direct instruction of word meaning and texts, oral reading, memorization, copying and repetition is very common (Zhou & Mcbride-Chang, 2009). Although word recognition is regarded as the foundation of learning to read, the role of letter sound knowledge and phonological awareness is not acknowledged in the national syllabus. Letter sounds and spelling rules are taught as knowledge of English phonology for the purpose of correcting pronunciation, and is often introduced in upper grades and adult English education (Ministry of Education P.R. China, 2011). The phonological-based instruction is not included in early grades, specifically for the purpose of keeping the instruction focused on language development, and sparing children the burden of tedious and out of context drills. However, in practical application reading activities are commonly adopted in early English education, as it is simple to implement, and this form of instruction suits the whole-class setting featuring teacher-centered lecturing and paper–pencil activities. Students are thus exposed to print material from the very onset of English education, and learn the English language by reading (Rixon, 2011).   36  Language Immersion Language immersion, originated in Canada in 1960s, is a foreign language teaching approach, where the approach aims to provide a communicative environment where the instruction emphasis is targeted towards interaction, conversation, and language use (Lightbown, Spada, Ranta & Rand, 2006). Moreover, the immersion educational intention is for students to use the target language authentically. To reach this goal, instruction of content knowledge, such as the subject of science, is performed through the target language. By integrating content learning with language learning, children are exposed to oral and written language of a greater variety and amount (Lightbown, Spada, Ranta & Rand, 2006). Students are immersed in the foreign language in the school setting so that they will compensate for the lack of exposure outside of school. Throughout this process, students’ knowledge of the local culture and skills in the first language should be activated while learning the foreign language. The immersion curriculum will perform in parallel to that of the local curriculum. Teachers are required to be bilingual so that they are able to compare and contrast languages, thus enabling the students to develop meta-linguistic awareness (Cummins, 1998, 2005, 2008). In Canada, immersion programs vary in specific combinations of two factors (Lotherington, 2004; Swain & Johnson, 1997): the initial age of immersion (i.e., early immersion beginning at kindergarten, middle immersion starting at age 9 or 10, and late immersion at the secondary level) and the amount of time allocated to immersion (i.e., partial immersion, allowing 50% of curriculum time for foreign language immersion throughout schooling, and total immersion, beginning with 100% immersion and gradually reducing the proportion to about 50% of time toward the end of junior schooling).   37  Two principles of pedagogy in immersion education are additive bilingualism and linguistic interdependence. Additive bilingualism is defined as students achieving proficiency in the target language without any detrimental effect on L1 competence, and that students gain cognitive and academic benefits of being a bilingual. According to the threshold theory (Cummins, 1979), one has to reach high oral proficiency in both languages to achieve additive bilingualism. Students in partial immersion have difficulty achieving high proficiency in French, thus the goal of additive bilingualism is compromised compared to students in total immersion (Cummins, 1998).  The theory also proposes that if students fail to achieve proficiency in either language, they will experience negative effects of learning two languages.   The second pedagogical principal of immersion education is linguistic interdependence, where students’ achievement in the target language largely depends on literacy and academic achievement in the first language. Language and literacy skills are transferable across two languages (Cummins, 1979). Language transfer is the reason that students in immersion programs performed equally when compared to students in non-immersion programs on first language proficiency and literacy. Transfer could occur bilaterally  from the first language to the target language and from the target language to the first language (Cummins, 1998).  In order to facilitate language transfer, students in immersion programs should be given ample support of literacy in L1 in and outside of the school setting. Students who have problems learning literacy in the target language should be provided with adequate  instruction on literacy in the first language to make an initial literacy breakthrough (Cummins, 1998, 2005). English Immersion in China English immersion programs in China commenced in 1997 for the purpose of reforming English education in China from the traditional grammar-translation approach to the  38  communicative approach. University professors, government officials, school principals and teachers from China, Canada and the United States formed a working group (CCUEI) to establish English immersion programs initially in approximately one dozen schools in three cities; Xi’an, Beijing and Guangdong. The English immersion is targeted primarily for early grades, from kindergarten to grade 3 (Qiang & Siegel, 2012; Siegel, 2011).  The initial concept as well as implementation processes of the English immersion program is a replication of French immersion programs from Canada. Modifications are made due to constraints of the local curriculum and limited resources. Hours of English immersion were reduced to 50% of the curriculum time, and subsequently qualified as partial immersion (Qiang & Siegel, 2012). Teachers in the immersion program are Chinese native speakers. Although they are certified teachers and have received intensive training to teach Immersion classes, they themselves still have limited English language proficiency, and many of the teachers cannot be considered to be bilingual. Large class sizes makes group activity and teachers’ individual interaction with students difficult (Qiang & Siegel, 2012; Song & Cheng, 2011). Despite all of these limitations, English instruction in immersion classrooms distinguishes itself from traditional English teaching in China. The primary goal in teaching immersion programs is oral language proficiency. Instruction focuses heavily on conversing skills and listening comprehension. A substantial amount of subject content is integrated into English instruction. In addition, teachers and students mostly speak English in immersion classes, whereas Chinese is dominantly used in traditional English classrooms (Siegel, 2011). The immersion program has a positive effect on attitude towards learning English, as children in the immersion program are more willing to communicate, and perceived higher  39  English competency.  No difference was found between the two groups in respect to motivation. Motivation in the study was only  measured by the effort of studying English (Knell & Chi, 2012). Considering the intense academic pressure and the competitive school environment in China, it is no surprise that students tend to put a lot effort on main school subjects such as English regardless of the nature of the program. However, students learning English in different ways may differ in other dimensions of motivation, such as ratio of intrinsic versus extrinsic motivation.  In terms of language and cognitive development, the English immersion program in China is an additive bilingualism program despite the fact it only qualifies as partial immersion (Cheng, 2012). Students in the immersion program performed significantly better than students in non-immersion programs on English oral language proficiency, English word reading, and English phonological awareness with no detriment to their Chinese language skills and academic achievement (Cheng, Li, Kirby, Qiang, & Wade-Woolley, 2010; Knell, 2010; Knell et al., 2007; Li, Kirby, Cheng, Wade-Woolley, & Qiang, 2012). No evidence is found showing the English immersion program has a detrimental effect on students’ Chinese language skills, except for one study (Li et al., 2012) where immersion students in grade 2 performed lower than students in the non-immersion program on Chinese phonological awareness, but the disadvantage disappeared in grade 4. Students in the English immersion program in China receive abundant instruction on Chinese literacy, and no less than students who are enrolled in the non-immersion program (Li et al., 2012). Cross-linguistic transfer of comprehensive language proficiency was observed among immersion students, and the correlation was stronger for students in higher grades with more experience. In contrast, the reverse pattern was observed for children in non-immersion programs  40  (Cheng et al., 2010).   Knell et al.'s (2007) study investigated cross-linguistic transfer of decoding abilities and found that English decoding and Chinese character reading were correlated for students in the immersion program. Knell (2010) enhanced this study, and recruited students across six grades from immersion and non-immersion programs. This study found that cross-linguistic transfer was a function of proficiency of both decoding and language skills. Pinyin reading was strongly correlated with English word reading, and the correlation was stronger for children who were in the low ability group. In contrast, the cross-linguistic correlation of reading comprehension was stronger for students in the high ability group. The influence of language immersion on cross-linguistic transfer was not investigated. The Matthew Effect in English Immersion Since English immersion has shown to be superior to regular English teaching in improving students’ English language and literacy skills, the Matthew effect was observed in terms of language learning in local schools(Knell, 2010). Gaps were created between students in the immersion and non-immersion program from the onset. A screening test was applied at school entry. The students who had prior experience of English learning were enrolled in the English immersion program, of which most were from households with greater social economical recourses (Knell, 2010).  The gaps widened across time, such that the difference between the two groups became larger after one school year regarding word reading, language proficiency and reading comprehension. The gap of word recognition increased across one year particularly for children in grade 3 and lower, whereas the group difference stabilized for children in grade4 and above. The oral language proficiency and reading comprehension gaps were found to have widened across the year for children in all grades.   41  According to Stanovich (1986), the widening gap of reading ability is attributed to the increasing difference of reading experience and print exposure. Enhancing students’ literacy skills such as word decoding and phonemic awareness early on may facilitate students’ ability to read independently and gain more reading experience, thus the Matthew effect can be alleviated. The theory was derived from the typical classroom environment in western societies where books of children literature are abundant, and students are encouraged to spend time on independent and interactive reading in as well as out of school.  Literacy education, notwithstanding if it is delivered in Chinese or English, in a Chinese context is very different from the western context. Literacy education in China focuses heavily on skill building while independent reading of books at school is not encouraged. In the current context of English immersion, the gap in reading experience and print exposure was artificial rather than the cause of disparities in reading abilities. Students in immersion programs received more reading material and intensive instruction, thus they gain more experience with English reading and writing, which leads to higher English literacy skills compared to their counterparts in non-immersion programs.  Direct instruction of basic English literacy skills to young students in non-immersion programs may compensate for their lack of reading experience at school in terms of skill building. Mastery of the basic skills may kick-start independent reading outside of school and prepare students for reading more difficult materials that will be introduced in subsequent schooling.  Theoretical Framework Constructs assessed in this research are English phonological awareness, English letter knowledge, English phonological decoding, English oral language proficiency, English literacy  42  abilities, Chinese literacy abilities, and at home resources. English phonological awareness was indicated through the measures; syllable awareness, rhyme detection, initial phoneme detection, final phoneme deletion and initial phoneme deletion. Indicators of English letter knowledge included letter naming and letter sound correspondence. English phonological decoding was assessed utilizing of non-word reading tasks. English oral language proficiency assessed two abilities, expressive vocabulary and syntactic awareness, which were indicated by picture naming and oral cloze respectively. English literacy abilities were assessed in this research through word reading and spelling. Chinese literacy abilities were assessed through Pinyin spelling and Chinese character identification. Influencing factors explored in this research included time, type of English program, whether to receive explicit instruction on phonological awareness. The confounding factor is students’ age and experience of learning English outside of school settings. According to the theory of orthographic mapping (Ehri, 1992, 2014) and lexical quality hypothesis (Perfetti & Hart, 2002), phonemic awareness has a causal influence on word recognition. Therefore, earlier English phonemic awareness should predict later performance on word recognition.  In addition, explicit instruction on phonemic awareness will have positive effects on phonological decoding and word recognition. The instructional effect should be maintained after the control group has learned alphabetic principles, as phonemic awareness makes a contribution to word recognition independent of letter sound knowledge. These two components should not be inter-changeable in instruction.  Moreover English oral language abilities should also predict performance of English word recognition. A basic level of vocabulary is the premises of learning to read. Syntactic awareness governs the process of predicting meaning of an unknown word during reading.     43  According to the lexical restructuring model, segmental representation of words are derived from expressive vocabulary (Metsala & Walley, 1998). Therefore, earlier levels of expressive vocabulary should predict later performance of phonemic awareness. Moreover, if explicit instruction on phonemic awareness is provided, the instructional gain on phonemic awareness should be predicted by the prior expressive vocabulary of students.  According to the interdependent hypothesis of bilingualism, students’ skills and knowledge of L1 literacy influence their performance in second language literacy (Cummins, 1979, 2005). Therefore, students’ earlier performance on Pinyin spelling and Chinese character identification may predict students’ later achievement in English reading and spelling. Concurrently, the transfer in the reverse direction is also possible. Instruction of English phonological awareness may also have positive effects on Chinese Pinyin knowledge, as the Pinyin writing system is very similar to the English writing system. The English immersion program provides intensive instruction which is content integrated and meaning centered. Thus, students in the immersion program have greater exposure to oral and written English. Therefore, they should have higher oral proficiency and reading abilities. Their underlying reading skills, including phonological decoding and phonemic awareness, are consequently higher, as suggested by the hypothesis that children learn to read by reading. Students with an enriched language and literacy environment will thus achieve superior results from decoding related instruction.  Moreover, according to the threshold hypothesis, cross-linguistic transfer of underlying skills only occurs when students reach a certain level of proficiency in both languages (Cummins, 1979). Students in the non-immersion program may not receive enough language exposure and  44  thus not reach the proficiency threshold for the cross-linguistic transfer. Therefore, the experience of English immersion may enable the cross-linguistic transfer.  Research Questions For children in primary grades who learn English as a foreign language in China, the following questions were asked, 1. Does the English immersion program have positive effects on English oral and literacy skills? 2. Do students from different English programs develop at the same rate for English literacy skills development?    3. Are concurrent and longitudinal predictions of word recognition from oral and phonological abilities the same across different English programs?   4. Does phonemic awareness have causal influences on word recognition? 5. Is phonological awareness training effective on English phonological awareness? Does the instructional effect transfer to English literacy outcomes and Chinese Pinyin spelling? If yes, are the instructional effects sustained after all the students, including students in the control group, learn alphabetic principles? 6. Do students’ prior Chinese literacy skills and English oral proficiency predict their gains from the English phonological awareness instruction? 7. Does phonological awareness predict students’ subsequent knowledge of letter sounds? Study 1 answered questions 1 through 4 by investigating the longitudinal prediction of English and Chinese skills to subsequent English literacy outcomes. It compared the students in two English programs on the growth of English literacy skills, and on the strength of the association between predictors and outcomes.  Study 2 answered questions 5 to 7 by  45  investigating efficacy of English phonological awareness instruction, and the delayed effect, one term later, after all students received instruction on alphabetic principles. Study 2 further investigated whether the instructional effects were moderated by students’ prior skills of Chinese literacy and English oral proficiency.    46  CHAPTER 3: STUDY 1 DEVELOPMENT OF ENGLISH READING SKILLS AMONG CHINESE EFL CHILDREN AND INFLUENCE OF LANGUAGE IMMERSION PROGRAM Introduction Since 1997, many schools in the city of Xi’an, China have adopted English immersion as an alternative approach to traditional English education, as this traditional approach had been criticized for being of the grammar-translation approach, and not appropriate for children. Studies have shown that the English immersion program in Xi’an lives up to the standard of additive bilingualism (Cheng et al., 2010; Qiang & Siegel, 2012). The program has been shown to have positive effects on students’ English oral language proficiency and literacy abilities without any detrimental effect on their Chinese oral language and literacy abilities. This program has faced many challenges, one of which is that schools and classes opted out of the program in higher grades, as the curriculum becomes more difficult, thus resulting in students having difficulties with reading and understanding the learning materials, which contains advanced vocabulary. Studies have shown that the immersion program, which focuses mainly on oral language development, have limited effects on underlying reading skills such as phonological awareness and word attack in early grades (Knell et al., 2007). To address this issue, phonics instruction, which focuses on teaching basic skills of English word decoding, was introduced in the immersion schools in 2010 (Chi, Lan, & Zhang, 2010). This instruction was also delivered to students in the non-immersion program. The skills tapped in phonics instruction are very important for reading in English. It has moderate to significant effects on English reading abilities among students in English-dominant countries (Ehri, Nunes, Stahl, & Willows,  47  2001). No study has been conducted to evaluate the effectiveness of the phonics instruction with students residing and being educated in Xi’an. Students in Xi’an receive English language education in two formats. These are English immersion and non-immersion programs. Compared to the non-immersion program, instruction in an immersion program is more intensive, and the curriculum materials cover a wider range of topics. And also, the instruction is fully mediated by English whereas Chinese is largely used as an aid in the non-immersion program. Studies show that English immersion has become a privilege in terms of education for students in the local community.  Students in the English immersion program are from households of higher socio-economoic status, and have more exposure to English outside of the school environment. The parents of children in the immersion program are more willing to invest in their child’s English education (Knell, 2010). This creates the Mathew effect, which was first identified in literacy learning in one’s native language (Stanovich, 1986), that the gap between poor and good readers grows larger and larger. The Mathew effect is relevant in the context of English language learning in Xi’an primary schools, as the gap between students in immersion and non-immersion programs widens across time.  This study compared students enrolled in the immersion and non-immersion programs on their growth of English oral and literacy skills, while receiving phonics instruction. The roles of the above noted predictors (phonological skills, oral language abilities and literacy skills in Chinese) of subsequent English word reading were also explored and compared across two programs. Research Questions  Does the English immersion program have net positive effects on English literacy skills?  48  Do the differences between the immersion and non-immersion program regarding English literacy outcomes increase or decrease after both student participant groups have received phonics instruction? Does the English immersion program influence the concurrent and longitudinal predictions of Chinese and English literacy skills regarding English word reading? Does prior phonological awareness have a causal influence on later word reading achievement among students in lower grades who learn English as a foreign language?   Method Participants A hundred and ten students (57 boys and 53 girls) from grade 2 (n = 64) and grade 3 (n = 46) in a large primary school (about 1,000 students) located in Xi’an, China participated in the study. Xi’an is the capital city of Shaanxi province in north western China, and has a population of 8.69 million. The average annual income of local residents was 30032 RMB (approximately 4700 USD) per person, ranked 16 among the 27 capital cities in China (Shaanxi Provincial Bureau of Statistics, 2016). The school is located in a newly developed neighborhood at the edge of the city. The children of factory workers, engineers and local tradesman attend the school, which serves as a mixed of lower-middle class neighborhood.  Data on educational level of the students’ mothers were obtained as an indicator of the social economic status of the household. 30.8% of the mothers did not proceed with education after the nine-year mandatory education, 26.7% had a secondary school degree, 30% possessed a college diploma, 10% possessed a bachelor degree and 2% had a master degree or above. Fifty-nine students were enrolled in the English immersion program (Mean age = 95.2 months). Fifty-one students were enrolled in the regular program (Mean age = 96.3 months). The  49  participants were in 10 classes. Five classes were a part of the English immersion program, and five classes were a part of the non-immersion program. The program membership was assigned upon students’ entry in grade 1 based on parents’ willingness and students’ previous exposure to English. A screening test was performed for program enrollment, and students who understood simple English (e.g. responded correctly to oral commands) were selected to enroll in the immersion program. Students selected for enrollment usually had received English lessons in kindergarten.  All student participants received English instruction from grade 1 onward. Five teachers participated the study. All of the participating teachers were Chinese native speakers and English major college graduates, with varying experience in teaching. The longest serving teacher had eight years of expereince, three had five years of experience, and the final teacher had one year. Each teacher taught two classes, one of which was the English immersion program, and the other one as part of the non-immersion program.  Non-immersion English Program  The student participants enrolled in the non-immersion English program were required to complete four 30 minute English lessons each week. Three of these lessons were dedicated to the government-endorsed curriculum, and one lesson was spent with a native English speaker reviewing the content. The curriculum covered themes including shapes, feelings, jobs, and seasons for grade 2 students, while the grade 3 students covered friendship, interesting opposites, special days, and hobbies. Each theme contained activities specifically targeted language outcomes including vocabulary, sentence patterns, pronunciation, and conversations.  English literacy education is described as the following. Text reading was not a component of the progam at this grade stage. Word recognition was taught by the “look-and-say”  50  method combined with incidental phonics. Copying words and sentences were largely involved in students’ daily exercise.  English Immersion Program Students in the English immersion program received nine English lessons a week.  Four lessons were identical to the lessons provided to the non-immersion students. The other five lessons were dedicated to the curriculum of the English immersion program. Two textbooks were used for the immersion curriculum, one of which focused on school subjects parallel to the local curricula including moral education and social science, and the other textbook focused on reading.  The social science curriculum contained themes including “China, The Country,” “Chinese History” and “China’s Achievements.” A typical lesson of social science included the following components; vocabulary instruction, introduction of the focus topic, teacher modeling of the topic in English, group activities such as fill in the blanks, creating charts, and sharing exercises in a group setting. A typical reading lesson would be composed of vocabulary instruction, story-telling using visual and auditory aids, text reading and answering comprehension questions.  As all the instruction was delivered in a class setting, individual interaction between the teacher and students was minimal.   Word recognition was embedded within vocabulary instruction, utilizing the following components, print form of the whole word, meaning in varying contexts, and phonological decoding and spelling practice. Reading words out loud and subsequently copying these words were assigned as homework.  Phonics Program At study commencement, the third week of the fall term in 2015, all the student participants started to receive synthetic phonics instruction, which was composed of two sessions  51  a week, with each session consisting of 20 minutes of instruction. The program consisted of two terms, for a duration of 26 weeks. The program consisted of both a primary and an advanced level. The primary level focused on the grapheme-phoneme -correspondence rules of all 26 Latin alphabetic letters, whereas the advanced level focused on 16 consonant clusters and digraphs. The program consisted of a total of 42 lessons, of which each was dedicated to one grapheme-phoneme-corresponding element. Teachers delivered 1 to 2 units in one session. Instructional components of the lessons included vocabulary learning, whole word recognition, decodable text, chanting nursery rhymes with the target element, recoding words into grapheme-phoneme units, analyzing the sound-print-correspondence element in words, and lower-upper case identification and conversion. See Appendix A for a sample lesson. During program implementation in the fall term 2015-2016, each teacher was observed for two sessions of instruction as a fidelity check. An instructional session was evaluated against the following fidelity criteria: (1) all key components of the lesson have been covered, (2) teacher has used the teaching materials in general accordance with the lesson plan, (3) there is no major distraction and/or disruptions during the lesson, and (4) the teacher has attempted to engage children’s attention in the lesson. Teachers followed the lesson plans, and no mistakes for the content in 8 of 9 sessions occurred. One session was interrupted by a fire alarm. After each observation, the researcher met with the teacher to discuss students’ response, to clarify the learning outcomes, and receive comments from teachers about the program. Structured feedback was given after each observational session on the instructional goals and teaching methods and techniques.   52  Instruments English language experience survey. Parents of participants were asked to complete a survey in Chinese to answer questions about their child’s home and extra-curricular English learning experience and home resources regarding English learning. The survey was self-developed with reference to Knell (2010). See Appendix B for details. The following variables were derived from the survey. Preschool English learning experience and extra-curricular English activities were binary variables; the response of “yes” was coded as 1 and “no” was coded as 0. Home English activity was the sum of frequency with which parents observed their child reading, speaking, singing and watching TV programs in English. Home resources was the sum of the numbers of items including books, magazines, DVDs, electronic devices and toys a child had at home regarding English learning. Parental involvement was the sum of frequency with which parents engaged in English related activities such as joint reading with their child. Parents’ education level was coded into five levels. Level 1 was grade 9 and below; level 2 was a diploma from a secondary school; level 3 was a diploma from a junior college; level 4 was a bachelor degree and level 5 was a master degree and above. Chinese character recognition. The Chinese Character Recognition Measure and Assessment Scale for Primary School Children (Wang & Tao, 1993) is a widely used test for screening Mandarin-speaking Chinese children with reading difficulties (Shu, Meng & Lai, 2003). In the application of this test instrument, the student participants were required to read aloud 141 characters, with the test halted if six consecutive items were answered incorrectly (See Appendix C). A total score of characters identified correctly was the result of this test instrument. The ordinal alpha instead of Cronbach alpha was employed to indicate the reliability of the data  53  in this study because the latter underestimated the internal consistency of ordinal item response data (Gadermann, Guhn, Zumbo, & Columbia, 2012). The ordinal alpha for this test was 0.99. Pinyin spelling. Students were asked to write down the orally presented syllables in Pinyin on a sheet of paper, which had the letters of the Pinyin alphabet printed at the top as a reference. These syllables correspond to multiple Chinese characteristics, with one of the corresponding characters presented to students as an aid. This test was administered in the group (See Appendix D). Students received one point for each correct spelling of the twenty four items of the test. The ordinal alpha for this test was 0.97. Oral cloze. This task was to measure the syntactical sensitivity of the student participant (Siegel & Ryan, 1988). Students heard a sentence with one word missing (e.g., Birds can––––.). The student was then asked to attempt to provide the missing word. They were required to provide a word that would create a semantically and syntactically well-formed sentence. The class of the missing word varied; nouns, adjectives, adverbs, prepositions, auxiliary verbs, and conjunctions were all used. Two demonstrative items and six test items were included (See Appendix E). The ordinal alpha was 0.72.  Picture naming (Learning Disabilities Association of Alberta, 2009). A picture naming task was adopted to assess children’s expressive vocabulary. Each child was presented with 15 pictures of common objects. Children were then prompted to name them one by one. Students scored one mark for every correct response. The test had 15 items (See Appendix F). The ordinal alpha for this test was 0.91. English phonemic awareness (Siegel & Ryan, 1988). Four tasks were employed to measure English phonemic awareness (See Appendix J). Pictures of words were presented to participants to reduce the load on working memory. Students received 1 point for every correct  54  response. The first task was English initial phoneme identification. Students were presented a target word orally and asked to identify the word that shared the same initial consonant with the target word from a list of three words. The task had six items. The ordinal alpha for this task was 0.97.  The second task was English initial phoneme deletion. Students were presented a stimulus word of one syllable and asked to sound out the remainder of the word after deleting the initial phoneme. The task had six items.  The ordinal alpha for this task was 0.91. The third task was English final phoneme deletion. Students were orally presented a word of one syllable and asked to sound out the remainder of the word after deleting the final phoneme. The task had 6 items. The ordinal alpha for this task was 0.92. The fourth task was English phoneme counting. Students were orally presented a word of one syllable and asked to count how many sounds the word contained. The items chosen are high-frequent words which were composed of 2 to 4 phonemes. The task had 6 items. This task was only used in time point three. The ordinal alpha for this test was 0.71. English rime detection (Muter et al., 1998). Participants were presented a stimulus word, and then were asked to choose the word that rhymed with the initial word from a list of three words (See Appendix I). The test administrator would read, "Here is a picture of a cat. Here are three more pictures". "Now which of these three, fish, hat, or sun, rhymes with cat?" The test question could be repeated by request of the student participant.  The child would receive 1 point for each correct word. The test included three demonstration items and ten test items. The task was implemented in previous research on the same population (Knell et al., 2007). The ordinal alpha for this test was 0.79.  55  English word reading. English word reading was measured by the word reading subtest of the Test of Word Reading Efficiency (TOWRE ; Rashotte, Torgesen & Wagner, 1999). The students were required to read aloud each word on the page. They received one point for each correct response. Considering the limited English proficiency of the participants, they were not timed on this task. The test had two demonstrating items, and thirty two test items (See Appendix G). The ordinal alpha for this test was 0.98.  English word reading silent. At time 3, children were assessed English word meaning access by the task of silent English word reading. Children were presented with a piece of paper with English and Chinese words on it. They were then required to choose the correct Chinese translation of each English word, from three options, to ensure that student participants understood the lexical meaning of the word. The test included fifty words. The ordinal alpha for this task was 0.89.  Pseudoword reading. The pseudoword reading subtest of the TOWRE test (Rashotte et al., 1999) was used to measure English decoding skills. Students were required to read pseudowords out loud, at their own pace, for which they received one point for each correct response. There were thirty-two test items and two demonstrative items in the subtest. Students were provided corrective feedback in demonstration items. The internal consistency coefficient for the test was 0.97. The items were further divided into two categories, the first of which contained nineteen one syllable words with regular spelling, based on CVC or VC structures. The second category contained thirteen words, composed of complicated spelling, containing digraphs, diphthongs, and consonant clusters. The rationale for the categorization was that the categories paralleled the focuses of instruction. The phonics instruction in term one was at a basic level, and focused on letter sounds and simple word decoding. The phonics program in  56  term two was at a more advanced level, and focused on grapheme-phoneme-corresponding rules of digraph, consonant clusters, and diphthongs (See Appendix H). Student participants were presented with a hardcopy containing all of the test words, for which one point was given for each correct response, where a correct response was defined as production with no obvious phoneme replacement, additions, or omissions. Procedure All students enrolled in grade two and grade three received the phonics instruction. Consent forms for assessment were send out to 450 parents. Of the 440 parent consents received, four declined participation. Of the remaining 436 students, 11 students were randomly selected from each class to take part in the assessment. The assessment was conducted, at three separate time points, across the 2015-2016 academic year. The time points of the assessment were at the start of the school year, September 2015, at the conclusion of the fall term, January 2016, and at the conclusion of the spring term, June 2016, respectively. The assessments performed at time one and time two were conducted in a quiet conference room within the school. The assessments for each student were of a thirty-minute duration. At time 3, the time allowed for assessments were reduced, as operational requirements of the school required a reduction of student availability. Individual testing was only feasible for tasks including word reading and pseudoword reading. Therefore, the assessment for group testing was adapted through the removal of expressive vocabulary, oral cloze, and Chinese character identification. Rhyme detection was adapted into a paper and pencil format. A test administrator read aloud a target word and three options, of which student participants were required to choose one word out of the given three which rhymed with the target picture. Two demonstrating items were presented to students before the test.  57  During program implementation, each teacher was observed for two sessions of instruction as a fidelity check. After each observation, the researcher met with the teacher to discuss students’ response, to clarify the learning outcomes, and receive comments from teachers about the program. Structured feedback was given after each observational session on the instructional goals and teaching methods and techniques. The lessons, two for each teacher, were observed to check for fidelity. Teachers followed the lesson plans, and no mistakes for the content in 8 of 9 sessions occurred.  Attrition One hundred and ten students participated in the assessment at time one, ninety-nine students participated at time two, and eighty-five students participated at time three. The sample size attrition was due to the tightened assessment schedule, as well as student absences. The Hawkins test yielded a non-significant result,  which indicated that there is not enough evidence to reject the assumption that the data is missing completely at random (MCAR) (Jamshidian, Jalal, & Jansen, 2014). Results Descriptive Statistics and Preliminary Analysis Descriptive statistics of participants’ home resources regarding English learning appear in Table 1. The amount of resources participants had at home regarding English learning was the same across two programs, except the number of books which was larger in the immersion students (F(1, 84) = 12. 65, p < .001).      58  Table 1  Means and standard deviations of number of items at home related to English learning  Immersion Non-immersion   Mean(SD) Mean(SD) F Books 3.74(7.81) 0.9(2.32) 12.65* Magazines 0.24(1.02) 0.22(1.41) 0.0025 DVDs 0.87(1.79) 0.65(1.37) 2.16 Electronic devices 0.92(1.19) 0.65(1.97) 1.46 Toys 0.78(1.16) 0.53(0.84) 3.58   Descriptive statistics of participants’ home English learning experience appear in Table 2. Around 20% of students in total had previous experience learning English before grade 1. A slightly larger number of students enrolled in the immersion program had learned some English before grade one, but the difference did not reach statistical significance (χ2 = 0.08, ns). Significantly larger number of students in the immersion program participated in extra-curricular English learning activities (χ2 = 13.63, p < .001). Parents reported that students in the immersion program spent more time on English-related activities at home including reading (p < .001), speaking (p < .001), and singing songs (p < .001). Parents of students in English immersion had higher education levels (both p < .001). The two groups did not differ on parental involvement (χ2 = 0.24, ns). Table 2  Percentages and medians on the background variables of students from two programs  Immersion Non-immersion χ2 Learned English before school (percentage) 20% 19% 0.08  Having extra curriculum English related activities(percentages) 21% 16% 13.63**  Father education(Max=5) 3 2 9.74* Mother education(Max=5) 3 1 15.05*  Child reading English(Max=5) 3 1 20.81** Child speaking English(Max=5) 5 4 11.91* Child singing English(Max=5) 3 1 17.91**  Child watching TV in English(Max=5) 1 1  Parents involvement (Max=5) 4 3 0.24  59  Extra-curricular English learning experience was the only background variable that showed significant correlation with the outcome variables. Therefore, it was selected to be the control variable in the following analyses. Means and standard deviations of the measures at three points appear in Table 3. Three MANCOVAs (Multivariate analysis of covariance) were conducted to compare the two programs on the outcomes at three time points, with age as the control variable. Effect sizes (Cohen’s d) were calculated using the F values (Lenhard & Lenhard, 2016). The statistics did not account for students clustering, thus they were indicative only. The two programs show significant and large differences in English word reading, silent word reading, picture naming, and oral cloze at all time points. The two programs are comparable in the variables including Chinese character identification, Pinyin spelling, letter naming, rhyme awareness, and initial phoneme identification, and final phoneme deletion at time 1, and show significant and moderate difference on pseudoword reading and initial phoneme deletion at time 1. At time 2, the two programs show significant difference in simple pseudoword decoding but not in complex pseudoword decoding. At time 3, two programs show significant differences in all the variables except rhyme detection.  Examination of the data revealed that considerable number of measures showed departures from normality, as indicated by the large values of skewness and kurtosis. The performance in final phoneme deletion and initial phoneme identification almost reached the ceiling effect at time one with an accuracy rate of 86% and 83%, and thus were removed from the following data analyses. Considering the distribution of the measures and the sample attrition, linear mixed-effects models estimated by the method of maximum likelihood were employed in this study. This statistical method does not require the data to be normally distributed and can  60  handle missing data by operating on covariance matrices. The analyses were conducted using raw scores in this study.  Development of English Word Reading Skills and Program Influence Linear mixed-effects models were employed in this study to account for the clustering of students. One model was built for each outcome measure. Time (three points), program (immersion and non-immersion), and interaction between time and program were set as fixed effects. The interaction term suggested a change in the difference between the two programs across time. The interaction term of zero indicated that the difference between the two programs did not change across two time points; a positive term indicated the difference becoming larger; a negative term indicated that the difference became smaller. Grade and extracurricular activities related to English was covariates and set as fixed effects too. Class, teacher, and individual student were set as random intercepts. The fixed effects for each outcome appear in Appendix K to R. In total, fifty p values of interest were produced. Benjamini-Hochberg (1995) procedure were employed to control false discovery rates. 61  Table 3  Means and standard deviations of all measures for students in both programs and both grades  Immersion  Non-immersion      G2 G3  G2 G3     Mean SD Mean SD  Mean SD Mean SD F p Cohen’s d Time1             PY spelling 21.32 3.10 21.54 2.11  20.62 3.32 20.00 3.39 1.96 ns  ChCharId 56.55 20.07 97.45 17.63  53.74 21.36 84.98 20.44 2.01 ns  PicNa 7.11 3.00 8.81 2.96  4.11 1.80 5.90 2.47 35.74 < .001 1.28 OClz 4.51 2.67 6.00 2.19  2.83 1.63 3.12 2.13 33.27 < .001 1.16 LeNa 19.64 6.17 21.00 4.07  19.74 5.15 20.38 3.47 0.38 ns  Wid 8.32 6.24 17.35 6.18  3.19 2.32 5.90 3.52 51.37 < .001 1.45 PsWid 1.81 2.18 5.08 3.32  1.00 1.27 2.57 1.89 11.94 .001 0.67 RhyA 5.34 2.32 6.32 1.30  5.56 1.60 4.90 1.97 1.73 .ns  InPhDe 2.04 2.22 4.29 2.09  1.57 1.95 2.95 2.01 4.37 .04 0.42 FiPhDe 5.13 0.95 5.52 0.68  5.15 0.82 4.85 1.35 0.25 ns  InPhId 5.21 1.61 5.00 1.36  4.71 1.15 4.71 1.12 2.19 ns  Time2             ChCharId 90.64 21.15 123.57 15.76  84.26 19.54 117.40 16.73 0.83 ns  PicNa 9.57 2.87 11.33 2.18  6.54 1.50 8.52 2.29 45.22 < .001 1.25 OrClz 5.86 2.62 7.85 2.05  2.55 2.00 4.15 1.79 41.73 < .001 1.32 LeNa 23.23 5.06 23.83 2.75  22.43 3.86 22.48 4.92 7.82 .007 0.57 WId 13.12 6.22 18.07 4.98  7.85 4.83 13.05 6.69 18.65 < .001 0.89 PsWid 4.50 2.95 6.15 4.05  2.92 1.93 4.71 4.08 8.02 .007 0.58 PsWidM  10.68 6.01 12.22 5.67  7.76 4.75 9.67 7.27 8.13 .006 0.59 SimPsWid 9.75 5.06 11.75 4.66  7.32 4.36 8.00 5.73 9.60 .003 0.64 ComPsWid 0.95 1.49 1.60 2.28  0.44 0.77 1.67 1.93 1.18 ns  RhyA 6.09 2.07 7.23 2.00  5.69 1.89 5.95 2.44 4.64 .04 0.44 InPhDe 3.09 2.48 5.65 2.16  2.19 2.13 4.00 2.17 3.55 ns  InPhId 5.20 1.61 5.66 0.79  4.88 1.80 4.84 1.30 6.15 .02 0.5  62   Immersion  Non-immersion     G2 G3  G2 G3     Mean SD Mean SD  Mean SD Mean SD F p Cohen’s d Time 3             PhCo 3.41 1.54 4.72 1.44  2.48 1.29 3.25 1.48 11.02 < .001 0.75 RhyA 5.65 1.92 8.4 1.18  5.12 1.92 7.75 1.35 1.8 ns  WId 15.24 5.96 23.90 3.79  10.68 4.62 16.10 6.63 27.75 < .001 1.18 WIdS 24.67 12.31 35.44 8.05  16.25 6.71 22.67 10.10 15.89 < .001 0.89 PsWIdM 11.91 6.96 21.38 4.83  7.88 5.40 14.60 7.24 5.16 .03 0.61 PsWId 8.59 6.73 17.57 4.91  5.11 3.81 11.35 6.36 4.97 .03 0.50 SimPsWid 10.38 4.40 16.07 2.27  6.72 4.62 12.50 4.92 5.22 .03 0.51 ConPsWid 1.57 3.37 5.54 3.04  0.76 0.83 2.10 3.23 3.86 .04 0.44 Note. PY Spelling, Pinyin Spelling; ChCharId, Chinese character identification PicNa, picture naming; OrClz, oral cloze; LeNa, letter naming; Wid, word identification; PsWid, pseudoword identification; PsWidM, word identification more items; SimPsWid, simple word decoding; ComPsWid, complex word decoding; RhyA, rhyme awareness ; InPhDe, initial phoneme deletion; InPhId, initial phoneme identification; FiPhDe, final phoneme deletion; PhCo, phoneme counting; ns, not significant. 63  Word reading.  All student participants improved in English word reading from time point one to time point two (Β = 5.75, p < .001), as well as from time point two to time point three  (Β = 3.00, p < .001). The students enrolled in the immersion program attained higher sores than students in the non-immersion program (Β = 5.02, p = .008).  Estimated means of word recognition of immersion and non-immersion students, after taking into account subject clustering and covariates, appear in Figure 1. The performance difference between the two groups significantly narrowed by time point two (Β = 2.34, p = .04). The trend of students in the non-immersion English program catching up with the immersion students in terms of word reading, however, was not sustained by time point three. The lines representing the two programs were parallel from time point 2 to time point 3.  Figure 1 Estimated means of word reading for two programs at three time points  Pseudoword reading. Both groups started with very limited English pseudowords reading skills. The average accuracy rate was 10% percent. Both groups improved significantly from time point one to time point two (Β = 1.95, p = .008), and from time point two to time point three (Β = 4.55, p < .001).  Students in the immersion program attained slightly greater  64  performance than students in the non-immersion program at time point one and time point two, but the group difference did not reach conventional levels of statistical significance (Β = 1.61, p = .07). As shown in Figure 2 A, student participants from the two programs started to diverge on pseudoword reading at time point three. The interaction effect between time and program at time 3, however, did not reach the conventional level of statistical significance (Β = 2.11, p = .06).  Figure 2 Estimated means of pseudoword reading for two programs at three time points  To further explore improvement made on decoding different types of pseudowords across time points, I divided the test items into two types. The first type of item was based on simple words, which were of a CVC/VC structure, and contained only regular spelling. The second  65  types of test items were of a greater complexity, and therefore more difficult to read, which contained complex patterns such as diphthongs, digraphs, and consonant clusters.   Simple word decoding. Decoding pseudo words of a simple structure improved significantly between time point two and time point three (Β = 2.35, p = .02), with no significant interaction effect between program and time (Β = -0.47, p = .70). A small difference was found between the two programs regarding decoding simple words (Β = 3.03, p = .03) favoring the immersion program, controlling for grade and outside of school English learning experience. Words with complex patterns. The two groups did not differ regarding decoding words of complicated patterns (Β = 0.23, p = .69) until time point three, as indicated by the significant interaction effect between program and time point three (Β = 1.46, p = .037). Students in the immersion program improved on decoding words with complex patterns from time point two to time point three, however student participants enrolled in the non-immersion program did not.  Outcomes that were not directly addressed in the program, including English phonological awareness, oral proficiency, and Chinese literacy skills, were also analyzed. Developmental trajectories on these outcomes served as a reference for the differences in literacy outcomes. Rhyme awareness. Rhyme awareness did not improve significantly from time point one to time point two (Β = .45, p = .25), nor from time point two to time point three (Β = 22, p = .40).  The program had a significant effect on rhyme detection (Β =1.04, p = .02), favoring the immersion program. Phonemic awareness. The immersion program had a positive significant effect on phonemic awareness (Β = 1.04, p = .04). Phonemic awareness improved slightly from time point  66  one to time point two for all the students, yet the time effect did not reach conventional levels of statistical significance (Β = 0.78, p = .06).  Oral cloze. The program had a significant effect on oral cloze (Β = 2.97, p < .001). The gap between the two programs did not increase at time point two, indicated by the non-significant interaction between time and program (Β = 0.83, p = .08).  Vocabulary. The program also had a significant effect on expressive vocabulary (Β = 2.52, p < .001). All students improved significantly from time point one to time point two (Β = 2.45, p = .008). The gap did not further diverge at time point two.  Chinese character reading. The program had no effect on Chinese character recognition (p > .05, ns), and the developmental trajectory did not differ between the two groups, indicated by the non-significant interaction term between time and program.  Concurrent and Longitudinal Prediction of English Word Reading from English Phonological Processing and Oral Proficiency Partial correlation coefficients among English outcomes at the three time points, controlling for age, are shown in Table 4. At time point one, word reading was not correlated with phonological awareness or with oral language proficiency for students in the non-immersion program. At time point two, phonological awareness was correlated with word reading. For students in the immersion program, oral proficiency and phonological awareness both were associated with word reading across two time points. Word reading and pseudoword reading were strongly correlated with each other for both groups of students across all three time points, and the association increased significantly at time two and three for students in the non-immersion program.   67  Table 4  Concurrent correlation coefficients among English variables at three time points   SWid Wid PsWid InPhDe RhyA PicNa OrClz Time 1        Wid NA  .66** .44** .48** .71** .59** PsWid NA .53**  .23 .37* .44* .39* InPhDe NA .15 .26  .38* .24* .37* RhyA NA .05 .35* -.07  .45** .38* PicNa NA .45* .44* -.10 .20  .74** OrClz NA .33* .37* -.04 .39* .45*  Time2        Wid NA  .68** .54** .50** .63** .59** PsWid NA .73**  .35* .46* .43* .57** InPhDe NA .42* .35*  .32* .40* .51* RhyA NA .22 .27 .07  .34** .53** PicNa NA .34* .18 .22 .01  .61** OrClz NA .08 .23 .10 -.15 .35*  Time 3        SWid  .38* .21 .06 .41* NA NA Wid .17  .71** .29 .07 NA NA PsWid .21 .73**  .27 .03 NA NA InPhDe .11 .03 .08  .13 NA NA RhyA .25 .50* .44* .05  NA NA Note. Coefficients for the immersion program are above the diameter; non-immersion coefficients are below the diameter. PicNa, picture naming; OrClz, oral cloze; SWid, silent word identification; Wid, word identification; PsWid, pseudoword identification; RhyA, rhyme awareness ; InPhDe, initial phoneme deletion; NA, not available.  Partial correlation coefficients of the relation between prior predictors with subsequent outcomes controlling for age appear in Table 5. Performance on word reading was stable across time points for the immersion students, as suggested by the strong correlation among measures at three points. The skill was developmentally less stable among the non-immersion students than the immersion students.  Word reading at time one was moderately correlated with word reading at time one (r = 0.42) and three (r = 0.44) among non-immersion students, however the association between time point two and three increased (r = 0.63). Development of pseudoword reading was less stable compared to word reading across time, but displayed a trend towards stabilizing (immersion: time 1&2 r = 0.33, time 2&3 r = 0.56; non-immersion: time 1&2 0.24,  68  time 2&3:  r = 0.53).  Most of the prior predictors were significantly associated with subsequent outcomes among the immersion students, whereas only a few correlations were significant among the non-immersion students. Table 5 Correlation of predictors at time 1 and 2 with outcomes at time 2 and 3 among the immersion and non-immersion students  Time2 Time3   Wid Wid SWid Immersion t1    PsWid .50** .67** .29* InPhDe .28* .38* .20 RhyA .48** .55** .17 PicNa .69** .62** .54** OrClz .57** .53** .49** Immersion t2    PsWid  .53** .31* InPhDe  .56* .50** RhyA  .39* .36* PicNa  .50* .38* OrClz  .61* .44** Non-immersion t1    PsWid .21 .51* .38* InPhDe .06 .06 .07 RhyA .14 .26 .28 PicNa .22 .38* .02 OrClz .06 .17 .08 Non-immersion t2    PsWid  .47* .48* InPhDe  .42* .36* RhyA  .51* .19 PicNa  .35* .06 OrClz  .18 .07 Note. PicNa, picture naming; OrClz, oral cloze; Wid, word identification; SWid, silent word reading; PsWid, pseudoword identification; RhyA, rhyme awareness ; InPhDe, initial phoneme deletion; *, p < .05; **, p < .001.  To investigate the concurrent prediction of phonological processing skills to English word reading, one linear mixed model was developed for each predictor for each time point.  Vocabulary, age, and outside-of-school experience were controlled variables, and entered into the model as fixed effects. Predictor, program, and interaction between predictor and program were also entered as fixed effects. Class was entered as a random intercept.  69  Table 6  Standardized fixed effects and significance in models of concurrent prediction of word decoding from phonological decoding and oral language skills at two time points  Word decoding t1 Word decoding t2  β p β p Model 1     PsWid 0.11 .39 0.59 <.001 PsWid*imm 0.29 .04 -0.16 .32 Model 2     InPhDe 0.008 .36 0.26 .05 InPhDe *imm 0.14 .26 -0.06 .74 Model 3     RhyA -0.01 .74 0.26 .05 RhyA*imm 0.45 .02 0.03 .96 Model 4     PicNa 0.22 .08 0.41 .02 PicNa *imm 0.51 .01 0.32 .34 Model 5     OrClz .18 .15 -0.05 .69 OrClz *imm 0.37 .04 .95 .005 Note. imm, immersion program;  PicNa, picture naming; OrClz, oral cloze; Wid, word identification; PsWid, pseudoword identification; RhyA, rhyme awareness ; InPhDe, initial phoneme deletion.  After the control variables including age, English oral vocabulary and extra-curricular activities were factored out of the analysis, pseudoword reading, phonemic awareness, and rhyme awareness did not have a significant influence on word reading concurrently at time point one. However, the interaction terms of program and rhyme awareness, and program and pseudoword reading were significant (p = .05, p = .02). The positive interaction effects indicated that the two predictors had a significant influence on word reading among students in the immersion program. At time point two, pseudoword reading, phonemic awareness, and rhyme awareness were shown to have significant influence on the concurrent performance of word reading regardless of program status. None of the interaction effects were significant at time point two. The models of concurrent predictions of oral abilities are also represented in Table 6. Fixed effects of expressive vocabulary and syntactic awareness on word reading were not  70  significant at time point one. However, the interaction effects of program & vocabulary, and program & syntactic awareness were significant (p < .01, p < .04). The positive interaction effects indicated that the two predictors had a significant positive influence on word reading among students in the immersion program. At time point two, expressive vocabulary was shown to have a significant influence on the concurrent performance of word reading regardless of program status. Syntactic awareness had no influence among non-immersion students, however had a significant influence among immersion students at time two, as indicated by the significant interaction term (p = .005). To investigate the longitudinal prediction of phonological processing skills to English word reading, two sets of linear mixed models were adopted, one for each time point. Vocabulary, age, and outside-of-school experience were controlled variables and entered as fixed effects. A predictor at previous time points, program, and interaction between predictor and program were entered also as fixed effects. Class was entered as a random effect. The models of longitudinal predictions of word reading skills appear in Table 7. Table 7  Standardized fixed effects and significance in models of longitudinal prediction of word reading from phonological processing skills.   Widt2 Widt3 WidSt3  β p β p β  p Time 1 predictors       Model 1        PsWid 0.008 .96 0.44 .01 0.20 .36 PsWid*imm 0.27 .21 -0.09 .61 -0.10 .66 Model 2       InPhDe 0.18 .15 0.07 .55 0.01 .94 InPhDe *imm -0.13 .44 0.04 .82 0.07 .67 Model 3       RhyA 0.13 .28 0.21 .09 0.18 .19 RhyA*imm 0.15 .56 -0.10 .68 -0.32 .29 Time 2 predictors       Model 3       PsWid   0.33 <.001 0.24 .11 PsWid*imm   -0.16 .39 0.01 .96  71   Widt2 Widt3 WidSt3  β p β p β p Model 4       InPhDe   0.26 .04 0.20 .19 InPhDe *imm   -0.24 .21 0.23 .31 Model 5       RhyA   0.30 .02 0.04 .77 RhyA*imm   -0.36 .26 0.48 .25 Note. t1, time 1; t2, time 2; t3, time3; imm, immersion program; Auto, autoregressor; Wid, word identification; WidS, word identification silent; PsWid, pseudoword identification; RhyA, rhyme awareness ; InPhDe, initial phoneme deletion.  The results indicated that none of the phonological processing variables at time point one significantly predicted word reading at time point two. Pseudoword reading at time point one significantly predicted word decoding at time point three. Phonological awareness variables at time point one did not have a significant influence on word reading at time point three. All the phonological processing skills at time point two significantly predicted word decoding at time point three, however did not predict silent word reading. None of the predictions were moderated by program status.  The models of longitudinal predictions of oral abilities are represented in Table 8. The overall fixed effects of vocabulary and syntactic awareness at time one on word reading at time two were not significant. The interaction term between vocabulary and program is both positive and significant, indicating oral abilities at time point one had a significant effect on word reading (pronunciation) at time point two only among student participants enrolled in the immersion program.  The effect, however, did not remain significant after the autoregressor was acounted for within the models. The overall fixed effect of time one vocabulary on word reading (pronunciation) at time three was significant, however, there was no significant effect regarding silent word reading. The interaction term on silent word reading was positive and significant. This indicates that expressive vocabulary had a longitudinal influence on word pronunciation  72  among all students, regardless of program status, whereas it only has a longitudinal impact on silent word reading among the immersion student participants. The overall effect of syntactic awareness was not significant on neither word decoding, nor silent word reading, however the interaction term was significant on silent word reading. The syntactic awareness at time point one is a significant predictor to silent word reading among students in the immersion program.  Table 8  Standardized fixed effects and significance in models of longitudinal prediction of word reading from oral language abilities   Widt2  Widt3  WidSt3   β p β p β P Model 1 t1       PicNa 0.21 .18 0.44 .01 -0.03 .86 PicNa *imm 0.59 .04 -0.09 .61 0.77 .01 Model 2       OrClz 0.02 .87 0.27 .55 0.01 .91 OrClz *imm 0.57 .03 0.1 .82 0.59 .02 Note. t1, time 1; t2, time 2; t3, time3; imm, immersion program; Auto, autoregressor; PicNa, picture naming; OrClz, oral cloze; Wid, word identification; WidS, word identification silent.  Longitudinal Relationship between English Phonological Awareness and Word Reading  In order to improve understanding of the causal relationship between phonological awareness and word reading, a cross-lagged correlation analysis was conducted. Two sets of mixed models were built for phonemic awareness and word reading, as well as rhyme awareness and word reading respectively. The autoregressor and the predictor at a previous time point were set as fixed effects. Class was set as the random effect. More specifically, these analyses were to establish how likely it is that phonological awareness at previous a time point can predict subsequent performance on English word reading compared with how much the earlier English word reading, in turn, predicted later performance on English phonological awareness. As indicated in Figure 3,  word reading at time one significantly predicted phonemic awareness at time two, after phonemic awareness at time one was controlled (β = .49, p < .001).  73  In contrast, the prediction in a reversed order was not significant (β = .02, p = .70). It is more likely that word reading caused phonemic awareness instead of the other way around. The pattern then changed between time two to time three, as the prediction of phonemic awareness at time two to word reading at time three (β = .56, p < .001) was stronger than the reverse order (β = .30, p < .05).  As indicated in Figure 4,  word reading at time one predicted rhyme detection at time two, after rhyme detection at time one was controlled (β = .20, p =.05). In contrast, the reverse prediction was not significant (β = .13, p = .25). This pattern indicated that prior word reading had larger influence on subsequent rhyme detection than the reverse. The pattern then changed from time two to time three, as the prediction of rhyme detection at time two to word reading at time three (β = .24, p = .03) was stronger than the reverse order (β = .03, p = .87).          74                      Figure 3  Cross-lagged correlation coefficients and standardized fixed effects between word reading and phonemic awareness      Word reading Phonemic awareness Word reading Word reading Phonemic awareness Phonemic awareness 0.72** 0.75* 0.24** 0.33* 0.02 0.49** 0.57** 0.30* 0.41** 0.53** 0.35** Time 1 Time 2 Time 3  75            Figure 4 Cross-lagged correlation coefficients (on the straight lines) and standardized fixed effects(on the cross) between word reading and rhyme awareness Word reading Word reading Word reading Rhyme detection Rhyme detection 0.71** 0.74** 0.40** 0.39** 0.13 0.20+ 0.24* 0.03 0.40** 0.43** 0.34** Time 1 Time 2 Time 3 Rhyme detection  76  Cross-Linguistic Transfer Partial correlations between Chinese literacy outcomes at time point one with English literacy outcomes at all three time points, controlling for age, appear in Table 9. Table 9  Partial correlation between Chinese and English literacy outcomes controlling for age  Time 1  Time2  Time3    Wid PsWid Wid PsWid Wid WidS PsWid Imm        PY spelling  .29* .18 .25 .03 .47* .39* .37* ChChaId .68** .52** .64** .46** .63** .42** .48* Non-imm        PY spelling  .08 .02 .01 .29 .01 .05 .10 ChChaId .29* .38* .12 .24 .28 .50** .38* Note. Wid, word identification; PsWid, pseudoword reading;  WidS, word reading silent; Imm, immersion program; Non-imm, non immersion program; PY Spelling, Pinyin Spelling; ChCharId, Chinese character identification. *, p < .05; **, p < .001.  The results of partial correlation analysis controlling for age showed that Chinese character identification at time point one was moderately associated with English reading across all three time points (time 1 r = 0.68, time 2 r = 0.65, time 3 r = 0. 63) among students enrolled in the immersion program. The associations were stronger than those between Pinyin spelling and English measures, as the correlations were low between Pinyin spelling and English word reading at three time points (time 1 r = 0.29, time 2 r = 0.25, time 3 r = 0. 47). The association between Chinese literacy skills and English word reading was stronger for students in the English immersion program than for students in the non-immersion program. For non-immersion students, Chinese character identification at time point one was less associated with English reading measures across the school year (time 1 r = 0.29, time 2 r = 0.22, time 3 r = 0. 28).  Pinyin spelling was not associated with any of the English measures at any time points.   77  To further investigate the influence of language experience on cross-linguistic transfer, several linear regression analyses were conducted. Linear mixed-effects models were employed to investigate concurrent predictions of word reading from Chinese character identification and Pinyin spelling. English word recognition and pseudoword reading were dependent variables. Chinese character identification, Pinyin spelling, age, outside-of-school learning experience, program membership, the interaction between program membership with Chinese character identification, and Pinyin spelling were entered as fixed effects. Class was set as a random intercept. Fixed effects in the model of the concurrent prediction of Chinese outcomes to English word reading at time point one are shown in Table 10.  Table 10  Concurrent prediction of English literacy outcomes from Chinese literacy outcomes at time 1 and 2 after controlling for age and English vocabulary   Widt1  PsWidt1  Widt2  PsWidt2   β p β p β p β p Excur 0.04 .94 0.03 .56 -0.12 .64 -0.22 .56 PicNa 0.36 <.001 0.26 .01 0.45 <.001 0.33 .04 Age 0.17 .56 0.18 .09 -0.11 .91 0.08 .07 Imm -0.63 .11 -0.53 .36 -0.56 .35 -0.59 .65 ChCharId 0.05 .61 0.12 .22 0.17 .27 0.03 .89 PY spelling 0.03 .51 0.03 .66 NA NA NA NA PY:Imm   0.39 .32 0.23 .65 NA NA NA NA Imm: ChCharId 0.76 <.001 0.51 .05 0.84 .04 0.79 .13 Note. t1, time 1; t2, time 2; Excur, extra-curriculum experience ; Wid, word identification; PsWid, pseudoword reading;  WidS, word reading silent ;Imm, immersion program; PY Spelling, Pinyin Spelling; ChCharId, Chinese character identification; NA, not available.  The fixed effect of Chinese character identification was not significant on word reading and pseudoword reading, whereas the interaction terms between Chinese character identification and program status was significant at time one (word reading: β = 0.76, p <.001; pseudo word reading: β = 0.51, p =.048 ). This suggested that Chinese character identification significantly  78  predicted word recognition and pseudoword reading at time one among students in the immersion program. This prediction did not prove to apply to children in the non-immersion program. The model of concurrent prediction at time point two showed that the interaction term between Chinese character identification and program status was still significant on real word reading (β = 0.84, p = .04), but not the one on pseudo word reading (β = 0.79, ns). A longitudinal prediction of Chinese literacy outcomes at time point one to English literacy outcomes at time point two and time point three were investigated. Two sets of linear mixed effect models, one for each predictor, were built for word and pseudoword reading at time point two and time point three. English word recognition and pseudo-word reading were dependent variables; outside of school experience and grade were control variables and set as fixed effects; Chinese literacy outcomes, program status, and the interaction were set as fixed effects; class was set as a random effect. Fixed effects in models of English word and pseudoword reading at the second and third time points are shown in Table 11. The results indicated that Chinese character identification at time point one significantly predicted later English word reading (β = 0.23, p = .03) and pseudoword reading (β = 0.36, p = .04) at time point two. Chinese character identification at time point one significantly predicted later word reading (β = 0.44, p = .04), pseudoword reading (β = 0.32, p = .03) and silent word reading (β = 0.44, p = .002) at time point three. The influence of Chinese character identification was not moderated by program status. Pinyin spelling at time point one significantly predicted pseudoword reading at time point two (β = 0.30, p = .04). Pinyin spelling was not a predictor of the other outcomes at either time point. The influence of Pinyin spelling was not moderated by program status.    79  Table 11  The models of longitudinal prediction of English literacy outcomes from Chinese skills at time 1  Widt2 PsWidt2 Widt3 PsWidt3 WidSt3  β p β p β p β p β p Model 1           ChCharId 0.29 .03 0.36 .04 0.44 .04 0.32 .03 0.44 .002 ChCharId *Imm 0.42 .15 0.18 .56 -0.09 .87 0.11 .73 0.01 .97 Model 2           Py 0.06 .67 0.30 .04 0.04 .70 -0.04 .77 0.12 .28 Py *Imm -0.07 .89 -1.05 .12 0.64 .21 0.84 .18 0.52 .37  Note. t2, time 2; t3, time3; imm, immersion program; auto, autoregressor; Wid, word identification; PsWid, pseudoword reading;  WidS, word reading silent; Imm, immersion program; PY, Pinyin Spelling; ChCharId, Chinese character identification; Auto, autoregressor.   Discussion This study investigated the development of English literacy skills across one school year among students from two different English learning programs. All student participants received phonics instruction for the first time in their education during this study year. Purposes of this study included 1) to compare two programs on the growth of different English literacy outcomes, 2) to investigate the concurrent and longitudinal prediction of English word reading from English phonological skills and Chinese literacy skills. I was interested to discover whether the phonics program could narrow the gap of English learning caused by the difference of students’ English learning experience. Effects of the Immersion Program on English Literacy Skills  This study indicated that the English immersion program had large effects on English  real word reading, expressive vocabulary, and oral cloze across the school year. In comparison, the abilities to delete the first phoneme and reading pseudowords were low for students in both programs, rendering the program effect minimal on phonemic awareness and pseudoword  80  reading. The finding of the contrastive profiles is consistent with the previous study which evaluated the effectiveness of the English immersion program in Xi’an (Knell et al., 2007), such that the program effect was more pronounced on language proficiency than on phonological processing skills, such as phonemic awareness and pseudoword reading, at early grades.  The students in the immersion program were better at English real word reading but showed a small advantage at reading pseudowords. Students from both programs initially registered poor performance on identifying pseudowords purely through phonological decoding and grapheme-phoneme-correspondence rules without the support of lexical cues. This then indicated that the English immersion program evidently prepared students to build the lexical route of word reading as proposed in the dual route hypothesis (Coltheart et al., 2001). However, the relatively large amount of language experience and the enriched environment provided by the English immersion program may not be significant enough for students to build the non-lexical route of word recognition, which requires phonologically analyzing words and applying letter-sound-corresponding rules. Growth of English Literacy Outcomes across Two Programs One of this study’s purposes was to test the hypothesis that explicit instruction on word decoding skills can alleviate the Matthew effect of word reading. As expected, students in the English immersion program were from families that had access to greater resources. Due to the screening test for program enrollment, students selected for the immersion program were those who had been previously exposed to English from parents or kindergarten. Students in the English immersion program received more English related homework as reported by the teachers. Thus they spent more time using English at home, particularly on activities including reading, speaking and singing.  The Matthew effect was indicative in this regard, as students who had  81  greater exposure to English before start of school were achieving greater experience with English language in school.  The pattern revealed in this study is of two phases. The first phase (from time point one to time point two) showed the opposite pattern of the Mathew effect, as the gap between the two groups regarding word reading narrowed. The second phase (from time point two to time point three) showed that the developmental trajectories of word reading of the two programs were parallel. The gap of word reading between the two programs narrowed at time point two, indicating that students in the non-immersion program, who were in a relatively deprived environment in terms of English exposure, made a greater improvement. Explicit instruction on English alphabetic and word decoding may contribute to the improvement. Basic levels of word decoding skills can initiate the self-teaching mechanism in independent reading (Share, 1995). The skill may enable the student participants to understand the textbooks and acquire greater vocabulary. The large gain made by ELL students who were at a disadvantage was also shown in D’Angiulli, Siegel, and Maggi (2004)‘s study.  In that study, at the end of grade five, ELL students showed no difference regarding word reading compared to English L1 students. This improvement was attributed to the intensive early literacy program ELL students received throughout their entire primary school experience. The trend of narrowing, however, was not sustained from time point two to time point three in this study, and students from the non-immersion program did not further advance on the immersion program participants on word reading. The explicit instruction of alphabetic principles students received in this study was an insignificant amount, only twenty to forty minutes each week. The shortened and limited program played a restricted role in helping students in the non-immersion program to further progress on the English literacy outcomes.  82  Students in the immersion program, on the other hand, received a more intensive and comprehensive English program, thus showed a greater improvement on word reading from time point two to time point three.  The gap on pseudoword reading (from time point two to time point three) widened, which is consistent with the Mathew effect: students in the immersion program benefited more from the explicit instruction on decoding. The results of the item analysis showed that the widening gap was mediated by the word items that contained complex spelling patterns. Students in the non-immersion program had minimal performance of reading pseudowords with complex spelling patterns, such as consonant clusters, diphthongs, and digraphs. This indicated that the effectiveness of the phonics instruction is in function of both language proficiency and instructional content. When learning primary phonics, which mainly focused on letter-sound knowledge and simple word decoding, all students improved to the same extent regardless of their program status. When learning the advanced phonics, which focused on complex spelling rules, students in the non-immersion program showed little improvement. This may be because the non-immersion students had significantly smaller vocabulary which plays an important role in identifying words with complex spelling patterns (Ouellette, 2006).  Students in the immersion program had greater oral language abilities, and more exposure to print, and were thus more likely to be exposed to words with complex spelling patterns. When learning spelling rules, they could draw from their prior experience and had more opportunity to practice the rules to master the newly learned skill. In contrast, students in the non-immersion program may only encounter the words within the phonics session, and were not familiar with the meaning of the words and had no opportunity to practice decoding complex words using authentic reading.  83  Concurrent and Longitudinal Prediction of Word Recognition Immersion and non-immersion students showed different patterns in respect to concurrent predictions of phonological skills and oral language abilities to word reading at the first time point. Phonological processing skills including pseudoword reading and rhyme awareness, and oral abilities including expressive vocabulary and syntactic awareness concurrently predicted word recognition among students in the immersion program, but none of the predictors significantly predicted word recognition among non-immersion students. The difference suggests that the language and literacy proficiency were too limited for students in the non-immersion program at time point one to produce any patterns regarding association.  The program difference regarding concurrent predictions disappeared at the second time point, including pseudoword reading, rhyme awareness, phonemic awareness, and vocabulary. The variables significantly predicted word reading for all students. This finding may be attributed to the fact that students from both programs received phonics instruction, as the instructional experience influenced the underlying mechanism of word reading. Explicit instruction of letter-sound correspondence and word decoding can enable students to attend to the intra-lexical structure of words and read words phonetically (Thompson, Cottrell, & Fletcher-Flinn, 1996; Ziegler & Goswami, 2005).  However, syntactic awareness was only important to silent word reading among the immersion students. This implied that students from the immersion program adopted the strategy of prediction to recognize words, which is in part governed by syntactic awareness. The results of the longitudinal prediction showed that phonological decoding, phonemic awareness and rhyme awareness at time point two predicted subsequent word reading after vocabulary, grade and outside-of-school English experience were controlled for, and the  84  predictive influence did not differ across the two programs. This finding added to the evidence that phonological awareness is an important skill for word reading among students who learn English as a foreign language (Hu & Schuele, 2015; Yeung & Chan, 2013).  In contrast, the longitudinal prediction of oral abilities to later silent word recognition differed across the programs. Expressive vocabulary and syntactic awareness significantly predicted silent word reading only among immersion students. This reflected that oral language abilities were associated with the understanding of meaning of words differently, across the programs, and implied the difference in acquiring word meaning in the different programs. Immersion students may learn the meaning of words through English language exposure, whereas non-immersion students learn word meaning through other approaches, e.g. Chinese language translation.  Cross-linguistic Transfer Findings in the concurrent prediction at two time points consistently indicated that cross-linguistic transfer is in function of the language learning environments. Students in the immersion program demonstrated a stronger association between Chinese and English decoding skills. The absence of cross-linguistic transfer of decoding skills among non-immersion students may be attributed to the non-immersion students not reaching the threshold of English proficiency to enable language transfer (Cummins, 1979; Chen, 2011). In contrast, students in the immersion program reached the threshold at time point one and displayed a pattern of cross-linguistic transfer of decoding skills at both time points.  The finding that the cross-linguistic transfer occurred for word decoding contradicts some previous findings, that if two scripts are orthographically distinct, the skills and knowledge in decoding one script cannot transfer to the other (Melby-Lervåg & Lervåg, 2011). The correlation  85  between decoding skills across languages found in the previous studies with Chinese-speaking English learners was argued to be caused by low proficiency in English (Bialystok et al., 2005; Gottardo et al., 2001). They argued that the children used visual strategy, which is dominant in reading Chinese characters, to read English script, because they lacked phonological decoding skills. However, the stronger cross-linguistic association found within the immersion students did not support this argument, as students in the immersion program had a higher proficiency in English, and greater decoding skills than the non-immersion students.  Moreover, Chinese students in this study did not read English words purely logo-graphically, as phonological decoding was involved in the reading process among this group of students. Students’ word reading performance was strongly associated with pseudoword reading. The association between silent word reading and pseudoword reading was also significant, although smaller (r = 0.34). This finding, together with findings with Chinese children in Hong Kong and Taiwan (Hu, 2013; Yeung & Chan, 2013), opposed the interpretation of the previous cross-linguistic studies which state that Chinese-speaking children do not apply the strategy of phonological decoding in reading English words (Bialystok et al., 2005; Muljani, Koda, & Moates, 1998).    Influence of Pinyin on English literacy outcomes was shown to be present but limited. Pinyin spelling significantly predicted pseudoword reading at time two. This finding was expected, as the tasks share similarities such that 1) both scripts are alphabetical, and 2) both tasks measured purely the decoding skill, and no lexical retrieval was required. This finding was consistent with the study which showed that Pinyin skill was correlated with English decoding (Wang, Perfetti, & Liu, 2005). However, the influence of Pinyin on English word decoding was limited. Pinyin spelling did not predict real word reading at either time points. It is hypothesized  86  that Pinyin was no longer the focus of instruction for the grade 2 and 3 students in this participant sample, and they were not explicitly taught to apply Pinyin spelling rules to sound out English words.   The transfer of decoding skills across two completely different languages found in this study could be enabled by two factors. The first factor is the shared school experience of learning the two subjects (English and Chinese), as the curricula in the immersion program support development of literacy in both languages. Also, the teaching method of both languages resembled each other to some extent. For example, teachers of both subjects would be assigned homework such as copying words and paragraphs, and reading aloud was emphasized in the instruction of both subjects.   The second factor could be the underlying proficiency commonly shared by two languages (Cummins, 2008). One component of the underlying proficiency is phonological processing. Chinese character identification is found to be associated with English phonemic awareness and pseudoword reading in this study. This finding was consistent with the previous studies, which indicated that the process of phonological decoding underlies word reading in both English and Chinese (Hu & Catts, 1998), and that the underlying phonological processing involving Chinese character identification can transfer to English word reading, especially pronunciation (Keung & Ho, 2009).   Study Limitations  Although I speculated that the improvement of decoding skills students made can be attributed to the phonics instruction, the effectiveness of the instruction cannot be concluded from this study, as the comparison groups were not included, in which phonics instruction was  87  not delivered. The possibility exists that the improvement of literacy skills was attributed to the daily instruction students received, instead of phonics instruction. The word reading tasks in the previous two time points did not measure lexical access. Pronunciation of print words was scored, which can be achieved through phonological decoding.  Reading vocabulary among the students was a subset of words students could pronounce, as indicated by the finding in time point three, that the number of words student participants know the meaning of was less than which they were able to read out loud.  In the EFL context, semantic information of a word is acquired simultaneously with, if not later than, formal information(spelling and pronunciation specification) for most of the words, due to instructional constraints (Jiang, 2000; Ye & Zhou, 2009).  According to Jiang’s stage theory of lexical acquisition (Jiang, 2000), formal information is not directly linked to semantic information at the initial stage of lexical acquisition. Therefore, print-and-pronunciation correspondence does not guarantee access to lexical-semantic information for EFL students. Conclusion Students from the two programs followed different developmental trajectories in regards to English word reading, assessed at three time points throughout the school year. The non-immersion students developed significantly faster than immersion students from time 1 to time 2 in real word identification, and then the two groups developed at the same rate from time 2 to time 3. The immersion students developed faster on pseudoword reading from time 2 to time 3. The group difference was mediated by the immersion group’s advantage in decoding pseudowords with complex spelling patterns. The program differences on other variables including English phonological awareness, oral abilities and Chinese character identification did not change across time.   88  The program also had an impact on the prediction of word meaning access. The immersion students’ performances in word meaning access were predicted by previous English oral language proficiency, but the prediction did not stand among the non-immersion students. In contrast, all three skills of phonological processing significantly predicted subsequent word decoding after controlling for English oral vocabulary, regardless of program status. Yet, none of the phonological processing skills significantly predicted word meaning access.       89  CHAPTER 4: STUDY2 EFFECTS OF ENGLISH PHONOLOGICAL AWARENESS INSTRUCTION WITH EFL CHILDREN IN XI’AN, CHINA  The previous study provided correlational evidence that phonological awareness have influences on English word reading and spelling. This study explored the causal influence of phonological awareness on English language and literacy skills by investigating the effectiveness of phonological awareness instruction (PA instruction). Concurrently, the roles participants’ prior Chinese literacy skills as well as English oral proficiency played in predicting their responses to the instruction were also investigated.  Research Questions and Hypotheses Is the PA instruction effective in improving EFL students’ English phonological awareness, as well as English and Chinese literacy skills? Are the instructional effects sustained after students received phonics instruction? Does students’ prior Chinese literacy skills and English oral proficiency predict later English reading skills?  Does phonological awareness, the outcome of PA instruction, predict students’ subsequent knowledge of letter sounds? The following hypotheses were tested. First, students who have received PA instruction will perform better than students in the control group on English rhyme awareness, English phonemic awareness, letter-sound knowledge, word reading and spelling, and the instructional effects are significant at both post and delayed posttest. Secondly, students’ prior English and Chinese literacy skills would predict their gains from the instruction.  90  Method Participants Participants in this study were from one school located in a prosperous residential area in the center of the city. The school was privately operated, however received subsidies from the local government. Parents were required to pay tuition, and the school selected students for enrollment. According to the principal, the school had greater funding, as well as more resources than an average school in the vicinity. Compared to study 1, parents in study 2 have higher educational levels.  Eight intact grade 1 classes, as well as their teachers, from an elementary school located in the city of Xi’an, China, participated in this study. All students and teachers were native Mandarin Chinese speakers. Four teachers of the grade one staff were specialized English teachers, whose background information is indicated in Table 12. Ninety students (47 boys and 43 girls), whose parents signed the consent form, approximately eleven students from of each class (45 to 48 students in each class), were randomly selected to take the assessment. Eighty-seven students participated in the pretest, Eighty-five students participated in the posttest, and seventy-six students participated in the delayed posttest.  Table 12   Teachers’ background information Teacher Gender Years of experience Credentials BLG Female 1 MA degree English  CLR Female 16 BA degree English LYL Female 3 BA degree English KXY Female 1 BA degree English  Design and Group Assignment This study adopted a quasi-experimental design. Eight classes and four teachers participated in the study. Students were randomly assigned to each class at school entry. Each  91  teacher was responsible for two classes, which were randomly assigned as either an experimental or control group for this study.  Research Context  English learning. Grade one students in this study received two English curricula, government endorsed curriculum and English immersion curriculum. English immersion curriculum was an English-mediated comprehensive curriculum incorporating the subject content of arts, moral education, and science (Qiang, Siegel, & Huang, 2008).  The curriculum was theme based, covering topics including going to school, my body, and my family. Through each topic, students were engaged in a variety of activities including drawing, colouring, playing games, making hand crafts, etc. The government endorsed curriculum focused on language learning. The curriculum covered themes including colours, animals, and body parts. Through each theme, activities were targeted towards language outcomes, including vocabulary, sentence patterns, pronunciation, and conversations. There was overlap between the two curricula in terms of topic and vocabulary. Teachers synthesized the two programs in their teaching. Explicit instruction on PA and alphabetic principles were not included in neither programs. Student participants would attend one to two English lessons each school day, each lesson being of thirty minute duration, for an approximate total of nine lessons a week. Proportionally, three classes were dedicated to learning the national English curriculum, and the other six classes were dedicated to the English immersion program.  All English instruction was mediated in English, as required by the policy of the English immersion program.  As students had little experience in English, and the teachers were not native speakers, this was a challenging task. Strategies such as total physical response (TPR) (Asher, 1969), as well as visual and auditory aids, were commonly adopted by teachers to create  92  comprehensible input. A typical lesson included the following components, 1) a warm-up through listening to an English song and dancing to the music, 2) a main activity which targeted content outcomes, 3) vocabulary activities, such as matching the pronunciation to the meaning and spelling, 4) language practice, such as a teacher modeling conversing in English and students imitating, 5) homework and review. Simple text and word reading out loud after the teacher’s lead was a common practice of teaching students to read. Spelling and writing were not required at this level.  Chinese literacy learning. Students in the study participated in ten Chinese lessons a week, of which each lesson was of thirty minutes’ duration.  The students spend the first six weeks of school learning a systematic Pinyin program developed by the Chinese language teachers from within the school (See the outline and a sample lesson of the program in Appendix Y). The program was fast paced, as most of the students had learned Pinyin in kindergarten. After this stage, a standardized text book was implemented, as required by the government prescribed curriculum. The instructional components included story reading, Chinese character recognition and copying, and word building. The instructional activities of word level skills included Pinyin decoding, stroke sequencing, character copying, word cloze and character structure awareness. Detailed descriptions of each activity can be found in Appendix U.  Procedures Prior to the commencement of the instructional program, teacher training was provided by the researcher. The training program consisted of a three-hour workshop and three demonstration lessons. During the initial portion of the workshop, the researcher explained the program structure, learning objectives, and characteristics of the activities. To assist the teacher participants to be better equipped with the linguistic skills required for running the sessions, the  93  researcher included content on English phonetics and phonological awareness. During the later portion of the workshop, teachers were given several opportunities to practice implementing the activities described in the program.   Following the workshop, three demonstration lessons were provided by the researcher. The purpose of the demonstration lessons was to demonstrate how the activities were conducted, and to assure the teachers that the activities were engaging to students, as well as to eliminate  concerns that students may not be ready to learn phonological awareness given they had little experience learning English. During the implementation of the program, the researcher held meetings with the teachers once every two weeks to discuss the children’s progress, to clarify the learning outcomes, and receive comments from the teacher participants about the program. Structured feedback was given after each observing session regarding the teaching methods and techniques.  The pretest was administered during the third week of the term, after the training workshops had been completed. The posttest was administered during the first week of January, 2016, immediately after the PA instruction had been completed. The delayed posttest was administered during the last week of May, 2016. The assessment allowed for thirty minutes of time to be given for each student, and was administered in a classroom.  There were two sets of tests, paper-pencil group tests, and oral individual tests. Ten students from the same class received test administration simultaneously. The student participants received the group test first, and then the individual tests individually. To exclude the inter-rater bias, an individual assessment was divided into three parts, each was administered by one tester for all students. When waiting for their turns, students were engaged in quite teacher supervised activities, including drawing, colouring, and reading.  94  Three graduate students from a local university were recruited to administer the tests. They each received three hours of administration training from the researcher, specifying how to perform the test administration. Experimental Instruction The training program was adapted from the Firm Foundations Program, which is a Kindergarten classroom program developed by teachers in the North Vancouver School District, of British Columbia, Canada (North Vancouver District, 2001). The goal of the program was to prepare students to learn English literacy through teaching phonological awareness and letter knowledge. Instructional components of the program included syllable awareness, rhyming, phoneme segmenting and blending, and letter-sound mastery. The program was activity-based with minimal lecture instruction involved.   The following adaptations were made to create the program for the experimental group in the present study. 1) Content regarding letter knowledge was removed, thus no paper-pencil tasks were involved. 2) Only activities that were feasible at a whole-class level were included.  Activities that required the use of complicated materials such as game boards were adapted to the use of projectors, body parts, gestures and word cards. Activities that required one-on-one interaction were removed.  3) The vocabulary was adjusted to the students’ exposure. Only words that appear in students’ text books were retained. Words used in the program are listed in Appendix V. The phonological awareness instruction was delivered in the fall term of grade one. The instructional sessions occupied the first teaching periods of each day. The instruction took place twice each week, and was each of a fifteen to twenty-minute duration. A typical session consisted of three parts, singing a song to warm up, introducing new words using flash cards and  95  movements, and three to four activities developing phonological awareness. Teachers would utilize a picture to prompt students to recall a previously introduced word and then say it out loud. Teachers would utilize a word in a sentence and then would explain the meaning using body gestures. Teachers would then ask students a question to validate comprehension, e.g. “which one of the pictures is an egg?” A phonological awareness activity targeted at one specific phonological skill, e.g. phoneme segmentation. For example, after introducing the word “egg”, a teacher would model segmenting the word into two phonemes by tapping her forearm and elbow while saying /e/, /g/. The teacher would demonstrate the action several times before asking students to repeat. The teacher would then repeat the action using another word. This activity typically consisted of three to five repetitions. A sample lesson plan is included in Appendix W.   The experimental instruction duration was ten weeks, containing a total of nineteen sessions. Each session lasted 15 to 20 minutes. The program schedule is shown in Table 13. The first session introduced the program, and familiarized students with the process that was to be repeated throughout the following sessions. Since students demonstrated mastery of syllable awareness in the pretest and the first session, only two sessions were dedicated to syllable awareness. Six further sessions were dedicated to onset-rime segmentation, and another nine sessions were dedicated to phonemic awareness. The final session was on reviewing the activities on rhyme detection and phonemic awareness. Table 13  Outline of the PA instruction Week Instructional goals 1 Listen for words in a sentence. Identify compound words. Familiarize students with the form and rules of the activities 2 Syllable identification segmentation 3 Rhyme detection  96  Week Instructional goals 4 Rhyme detection and initial phoneme segmentation 5 Rhyme detection and initial phoneme segmentation 6 Phonemic awareness(phoneme discrimination and segmentation)-words of two or three phonemes 7 Phonemic awareness(phoneme segmentation and blending)-words of two or three phonemes 8 Phonemic awareness(phoneme discrimination and segmentation)-words of four or five phonemes 9 Phonemic awareness(phoneme segmentation and blending)-words of four or five phonemes 10 Review  rhyme detection, phoneme segmentation and phoneme blending  Fidelity Each class in the experimental group was observed twice, and evaluated against the following fidelity criteria:  (1) All key components of the lesson have been covered, (2) teacher has used the teaching materials in general accordance with the lesson plan, (3) there is no major distraction and/or disruptions during the lesson, and (4) the teacher has attempted to engage children’s attention in the lesson.  Eight sessions were observed in total, seven of which followed the lesson plan, meaning that the key instructional components were covered and the teaching materials were used correctly to address the lesson purpose. Teachers in all sessions made attempts to engage students, and no disruptions were observed.  One session in each control class (four in total) was observed to confirm that no PA skills were introduced during the period of the study.  97  Control Condition Classes in the control group were taught by the same teachers in the experimental condition. The teachers were instructed in advance that the status quo in the control classes must be maintained, and no instruction on phonological skills must be included in the lessons for the duration of the study. Vocabulary introduced in the experimental classes were also taught in the control classes.  The morning class sessions, with a duration of twenty minutes each, were dedicated to activities including language games (for example, Simon says, touch body parts), singing songs, and reading out loud while being led by the teacher.   Phonics Program after the PA instruction After the experiment classes receiving the PA instruction in the fall term, students from both the experiment and the control group received phonics instruction throughout the following spring term. The phonics program utilized was the identical program of primary phonics adopted for study 1. The instruction was delivered once a week for thirteen weeks. Each session was of a forty-minute duration, and covered two lessons of the program.  Instruments English language experience survey. Identical to study 1. Chinese character recognition.  Identical to study 1. Pinyin spelling. Identical to study 1. English oral cloze. Identical to study 1. English picture naming. Identical to study 1. English phonemic awareness. Identical to study 1. English rime detection. Identical to study 1.  98  English word identification. English word reading was measured utilizing the word-reading subtest of the Test of Word Reading Efficiency (TOWRE ; Rashotte, Torgesen & Wagner, 1999 ). The students were required to read aloud each letter or word on the page, and then received 1 point for each correct response. Considering the limited English proficiency of the participants, the students were not timed on this task. 80% of words (25) were simple words, and appeared in the students’ textbook. The ordinal alpha for this test was 0.94. Pseudo word reading. The pseudo word reading subtest of the TOWRE Test (Rashotte et al., 1999) was used to measure English decoding skills. Students were required to read pseudowords out loud, at their own pace, and were awarded one point for each correct response. There were twenty-one test items and two demonstrative items. Students were provided correct feedback in demonstration items. Spelling of all items complied with regular grapheme-phoneme-corresponding rules. The ordinal alpha for this test was 0.90. Invented spelling. Students were asked to write down orally presented words. All children were encouraged to do their best to write down the letters they thought most accurately represented what they had heard. The six words presented were all of simple CVC, CCVC and CVCC structures. The spelling of each word was rated using the scale in Appendix X, which was adapted from the criteria developed by Tangel and Blachman (1992). Data Analysis Linear mixed-models (Verbeke & Molenberghs, 2009) were employed to analyze the influence of time, instructional group, student, and teacher on English and Chinese language outcomes. The model parameters were of two categories, fixed and random effects. Fixed effects presumed that the effects are the same across observations, and the exact amount of the effects were estimated by the model. Random effects have two categories, random intercept and random  99  slope. If a random intercept is set for a factor, it indicated that the dependent variable varies randomly for said factor. A random slope set for factor A on factor B indicates that the effect of factor A on a dependent variable varies randomly across all the levels of factor B. The variance of the random effects rather than the exact amount is estimated by the model.   The linear mixed-model was adopted due to the hierarchical structure of the data, as presented in Figure 5. The structure can be conceptualized as containing a three-level hierarchy. Three time points were nested within students, which were nested in instructional groups, which are nested in teachers. Teacher factor was taken into account in the analysis for the following reasons. 1) It allows accurate estimation of instructional effects size, as the unit of group designation in this study is class not students. Using students as a unit of analysis and ignoring the class factor would result in an inflated instructional effect (Gersten et al., 2009),  2) the analysis took into account dependence among outcomes of students who were in the same class and taught by the same teacher. This resulted in more accurate estimation of standard errors.  To answer the first and the second research question, whether the instruction is effective immediately and in long term, one model was built for each outcome. Random intercepts were set for each subject and teacher, as the outcome variables were expected to vary randomly across students and classes. As the participating teachers were of different experience and styles, random slope of group was set for each teacher, indicating that the instructional effect might differ across teachers.  Group (experimental vs control), time (pre, post and delayed post) and the interaction effects between the two factors were of interest and set the fixed effects; prior experience of English learning (yes or no) was the covariate, and set as a fixed effect. Effect size at class level was calculated using the following formula,   100  d= ei/sqrt( varsub/ns+vartch*nc+varg*ng+ r/i*ns) ei , estimated parameter of group effect at time I; varsub ,variance across students; ns ,number of students per class;  vartch, variance of teacher;  nc, number of classes per teacher; varg , variance across group; ng number of classes per group; re, residual; i, times of assessment.  The instructional effect was indicated by the interactional term between group and time. The immediate effect was indicated by the advantage of the experimental group over the control group at time two, after controlling for initial differences between groups at time one, students’ prior experience of learning English, and gains of both groups made at time two.  The delayed instructional effect was indicated by the advantage of the experimental group at time point three, after controlling for the above factors. The significance of a fixed effect was obtained by comparing the current model with a reduced model, without the parameter of the fixed effect. It was hypothesized that the time and group effects, as well as the interaction effects would  indicate as positive, therefore, one tailed test was performed. The R (R Team, 2013) statistical analysis program and the lme4 package (Bates, Maechler, Bolker, & Walker, 2015) were employed to perform the analysis. The fixed effects were presented in the main text.   101   Figure 5 Data structure in study 2  Results and Discussion Background Information  Parents of eighty three students (93% of the entire sample) completed the questionnaire. Table 14 shows the demographic characteristics of participants. The test group and the control group did not differ regarding background variables, including home English learning activities, home resources related to English learning, and parents’ education level (all p > .05). Records on whether students had prior English educational experience before entering primary school were also collected. Six out of 45 students in the group had been exposed to English instruction in preschool, and fifteen out of 45 students in the experimental group had received English instruction in preschool. There was no significant difference between the two groups (Χ2 = 2.09, p = .09).   102  Table 14  Demographic characteristics of participants before program implementation Background variables   Experimental (n = 41) Control (n = 39)     Max Mean SD Mean SD  F English learning related activities at home 30 13.16 7.09 11.08 7.6 1.29 English resources at home 10 1.66 2.11 1.69 2.07 0.25 Mother education level 5 3.18 0.76 2.92 1.01 1.75  Primary Analysis  Means and standard deviations of students’ performance on all outcome measures at the three data collection time points appear in Table 15, together with the results of three MANOVAs regarding group differences at three time points. The values did not account for students clustering, hence were only indicative. A ceiling effect was observed for the final phoneme deletion measures at all time points (83 % students attained full marks), and initial phoneme identification at time two and three (accuracy rate of 85%). These two measures were removed for further analysis.  The results of the MANOVA at pretest revealed that there was a significant main effect of group (Wilks’ Lambda = 0.75, F (10, 83) = 2.50, p < .05). The results of follow-up ANOVA analyses indicated that there were no effects of group on all measures except on initial phoneme deletion (F (1, 89) = 14.89, p = .001), indicating that the experimental group performed at a greater level than students in the control group on phonemic awareness at pretest. The group inequality of phonemic awareness at pretest could be due to the fact that students in the experimental group had a head start advantage, due to having received three mini lessons prior to the pretest.    103  Table 15  Means and standard deviations of the outcomes across teachers at three time points   Experimental Control    Max Mean(SD) Mean(SD) F p Pretest      PY 20 17.60(3.10) 16.60(3.68) 1.84 ns ChCIde 89 33.45(19.10) 35.96(2.34) 0.37 ns Voc 15 4.20(3.43) 3.01(2.07) 3.81 .05 Letter Id 26 4.78(7.19) 2.64(5.49) 2.53 ns Word Id 32 5.33(6.00) 3.87(3.95) 2.14 ns PSW Id 21 3.98(4.30) 3.18(3.09) 1.25 ns RhyDe 10 4.90(2.20) 4.73(2.16) 0.10 ns OCl 8 2.20(1.71) 2.03(1.33) 0.25 ns IPD 7 1.73(2.23) 0.37(0.72) 14.89 .001 FPD 6 5.14(0.88) 4.77(0.82) 3.73 ns IPI 6 4.85(1.08) 4.53(1.29) 2.33 ns Posttest      PY 20 19.17(1.26) 18.23(1.69) 6.27 .01 ChCIde 89 46.13(18.90) 46.20(20.50) 0.03 ns Voc 15 6.18(3.15) 4.41(2.14) 7.47 .007 Letter Id 26 6.65(8.23) 3.77(6.06) 3.66 ns Word Id 32 12.07(7.66) 9.20(6.26) 7.34 .008 PSW Id 21 6.78(5.49) 3.22(2.96) 11.56 .001 RhyDe 10 5.83(1.71) 4.86(1.61) 11.52 .001 OCl 8 2.84(1.87) 2.04(1.60) 3.11 ns IPD 7 3.08(2.28) 0.82(1.28) 24.33 <.001 FPD 6 5.45(0.50) 5.15(0.47) 6.11 .02 IPI 6 5.27(0.96) 5.11(1.08) 0.35 ns Spelling 33 12.42(7.31) 9.11(5.27) 6.69 .01              104    Experimental Control    Max Mean(SD) Mean(SD) F p Delayed post test      ChCIde 89 61.59(13.50) 63.32(14.10) 0.29 ns Voc 15 10.26(2.25) 9.95(2.05) 1.59 ns Letter Id 26 12.84(8.45) 10.73(7.78) 1.29 ns Letter SK 26 15.46(5.73) 12.68(6.00) 9.11 .003 Word Id 32 23.80(7.26) 21.95(6.9) 4.80 .03 PSW Id 21 9.51(6.98) 8.35(5.92) 0.49 ns RhyDe 10 6.87(1.61) 7.27(1.57) 0.36 ns OCl 8 5.27(1.77) 4.73(2.12) 4.22 .04 IPD 7 3.92(2.38) 2.83(2.27) 3.73 ns FPD 6 5.36(0.53) 5.21(0.47) 0.75 ns IPI 6 5.02(0.99) 5.19(1.13) 0.46 ns Spelling 33 18.85(5.97) 13.09(5.70) 26.04 <.001 Note. PY, Pinyin spelling; ChCIde, Chinese character identification; Voc, vocabulary(picture naming); Letter id, letter naming; Letter SK, letter sound knowledge;   Rhyde, rhyme detection; OCl, oral cloze; IPD, initial phoneme deletion; FPD, final phoneme deletion; IPI, initial phoneme identification; ns, not significant. 105   Instructional Effects A full model of each outcome contained three parts, random effects, fixed effects, and residual terms. Random effects accounted for the individual differences across subjects, differences across teachers, and differences across classes taught by the same teacher. Fixed effects showed the estimated effects of time, group and the interaction between the two. Residual terms indicated that other factors affected performance outcomes that are beyond the scope of this study. Estimated means of all outcomes accounting for clustering and covariates at all three time points are plotted in Figure 6 .   106   Figure 6 Estimated means and confidence intervals of instructional outcomes at three time points controlling for English learning experience and teacher effect   107   Figure 6 Estimated means and confidence intervals of instructional outcomes at three time points controlling for English learning experience and teacher effect continued  Word recognition. Fixed effects of the linear mixed model of word recognition are shown in Table 16. The fixed effects of time two and time three were significant (p <.001 for both time points). It indicated that the students’ performance of word recognition improved significantly at both time points. The interaction terms between group and time was not significant at either posttest, or delayed posttest, indicating that the instructional effects were not significant at either time point.     108  Table 16  Fixed effects in the model predicting word recognition.  Β SE df t p Intercept 5.76 1.83 11.93 3.14 .005 Time 2 5.51 0.97 170.00 5.66 <.001 Time 3 17.40 0.97 170.00 17.86 <.001 Exp group 1.07 1.44 7.32 0.75 .24 No PEXLE -3.19 1.39 79.80 -2.29 .01 Exp group t2  2.08 1.37 170.00 1.52 .06 Exp group t3 0.90 1.37 170.00 0.66 .25 Note. No PEXLE, No priori experience of learning English; Exp, experiment   The significant improvement students attained in word reading may be attributed to the English instruction they received at school, as the daily instruction of English focused significantly on both vocabulary and word recognition.  However, students who received the PA instruction did not show to attain an advantage regarding word identification. Previous studies have shown that PA instruction without using letter and print had significant but small effects on word reading (Ehri, Nunes, Willows, et al., 2001). The non-significant finding in the present study may be due to the following reasons. Firstly, transferring phonological skills to word reading and spelling is difficult. Chinese children are more likely to use visual strategies to decode English words than their English speaking counterparts (Wang et al., 2005), as influenced by their native language. Thus applying the PA skill to word identification requires explicit instruction. Secondly, the quality of the instruction may be compromised by the shortened length of the instruction and the large group size, as the students in the classroom did not have ample opportunity to practice PA skills, thus were unable to apply them in real word reading.  Pseudoword reading. Fixed effects in the model of pseudoword reading are presented in Table 17. The delayed posttest parameter was significant (p = .01), indicating that all students in the experimental group improved significantly in pseudo word reading at delayed posttest. The significant parameter of experimental group at posttest indicated that the immediate instructional  109  effect was significant (p < .02, Cohen’s d = 0.77). However, the instructional effect was not sustained when the two groups were compared at the delayed posttest, indicated by the non-significant parameters of the group, and the interaction term at delayed posttest.    The estimated and observed group effects at the three time points for each teacher are shown in Figure 7. The observed patterns were convergent with the estimated patterns for three of the four participating teachers. The instructional effect was not evident for teacher four.  Table 17  Fixed effects of pseudo word reading  Β SE df t p Intercept 3.32 1.40 10.66 2.37 .002 Time 2 0.14 0.87 170.00 0.16 .43 Time 3 4.70 0.87 170.00 5.38 .01 Exp group 0.63 1.15 7.78 0.55 .31 No prior EXLE -1.02 1.03 79.51 -0.99 .18 Exp group t2 2.54 1.23 170.00 2.07 .02 Exp group t3 0.44 1.23 170.00 0.36 .34 Note. No PEXLE, No priori experience of learning English; Exp, experiment  A further result of the study, that PA instruction was effective on English pseudoword reading, a skill which was not directly addressed in the PA instruction, provides evidence for the theory that PA skills together with letter sound knowledge are essential for orthographic mapping, which was measured by pseudoword reading (Ehri, 1987). Although students had not formally received instruction in mapping of letters to sounds by time interval two, they have received Pinyin instruction, and had been exposed to English print.  Thus, they may draw from knowledge of Pinyin to derive sounds that the letters make in English words. Explicit instruction of English PA skills facilitated the process of transferring skills and knowledge of Pinyin to English word decoding. A follow-up analysis was conducted, investigating the moderating effect of prior Chinese and English knowledge on the instructional effect of pseudoword reading. The results and the discussion appear later in this chapter.  110    Figure 7 Instructional effect on pseudo word reading by teacher Pre Exp, predicted means in the experimental group; Pre Con, predicted means in the control group; Obs Exp, observed means in the experimental group; Obs Con, observed means in the control group.  Initial phoneme deletion. The fixed effects estimated in the model of initial phoneme deletion are shown in Table 18. The experimental group parameter was significant, indicating that the experimental group attained greater performance than the control group at all three time points (p = .005). The interaction term at the immediate posttest was significant, indicating phonemic awareness improved significantly at the immediate posttest among students in the experimental group (p = .025), however not within the control group. Both groups improved significantly in initial phoneme deletion at the delayed posttest, as indicated by the significant parameter of time 3 (p = .005). The immediate instructional effect was significant with an effect  111  size of 0.59. The developmental pattern of two classes taught by each teacher is shown in Figure 8. Observed data of three out of four teachers were convergent with the predicted patterns.  Table 18  Fixed effects in model predicting initial phoneme deletion  Β SE df t p Intercept 0.91 0.46 55.96 1.96 .025 Time 2 0.45 0.36 94.00 1.22 .11 Time 3 2.27 0.55 5.22 4.12 .005 Exp group 1.11 0.44 206.17 2.52 .005 No prior EXLE -0.61 0.38 83.67 -1.61 .055 Exp group at t2 0.99 0.51 168.41 1.94 .025 Exp group at t3 0.10 0.51 168.50 0.20 .42 Note. No PEXLE , No priori experience of learning English; Exp, experiment  The experimental and the control groups were not equal at pretest for the outcome of initial phoneme deletion, as the experimental group improved more than the control group on the outcome at the posttest. This pattern suggests there is a possibility for selection bias, specifically maturation bias such that the two groups may be maturing in different rates. However, this possibility can be ruled out, as the inequality was very specific on this outcome.  It is very likely that the specific inequality was due to the mini lessons given to the experimental group prior to the pretest.   Students in this study showed difficulty deleting the initial phoneme of a word, when compared to the greater performances attained on other tasks of phonemic awareness. At pretest, students presented with the ceiling effect for initial phoneme identification and final phoneme deletion, but presented with almost the floor effect on initial phoneme deletion. This finding is convergent with that in previous studies, that it is particularly difficult for Chinese children who speak either Cantonese (Huang & Hanley, 1995) or Mandarin (Tao, Huang, & Li, 2005) to delete the first single consonant from the word to break the CV unit (e.g., deleting /s/ from sit).This may be due to the constraint imposed by the students’ first language. Mandarin speech is  112  composed of simple syllables which are structured as consonant and vowel (CV). Boundaries between syllables are always at a position of vowel or nasal consonant. Therefore, deleting a consonant in a CV unit, for Chinese EFL students, is unique to the English language and is challenging.  Nevertheless, it is an important skill as shown in this study, as the skill was associated with students’ later performance of English literacy.    Figure 8 Instructional effect on initial phoneme deletion by teacher   113  Rhyme detection. Fixed effects in the model of rhyme awareness appear in Table 19. Neither group showed significant improvement in detecting rhyme awareness at posttest, as indicated by the non-significant parameter of time two (p = .24). In contrast, the performance of rhyme detection improved at the delayed posttest. The parameter of time three was significant (p < .001).  The instructional effect at either time points was not significant. Table 19   Fixed effects in the model of rhyme awareness  Β SE df t p Intercept 4.48 0.70 4.68 6.41 < .001 Time 2 0.26 0.36 170.00 0.70 .24 Time 3 2.51 0.36 170.00 6.90 < .001 Exp group 0.10 0.63 4.91 0.16 .44 No prior EXLE 0.04 0.79 2.95 0.05 .43 Exp group at time 2 0.59 0.51 170.00 1.14 .13 Exp group at time 3 -0.47 0.51 170.00 -0.91 .18 Note. No PEXLE , No priori experience of learning English; Exp, experiment  PA instruction was shown to be not effective in improving rhyme awareness, as the teacher participants reported that it was difficult to explain the definition of a rhyming pair to students. They said that the lack of ability was due to students simply not having a vocabulary with enough breadth to know enough words that rhyme with each other. When explained that rhyming words have the same ending, students only attended to the ending consonant, but were not able to perceive the medial vowel and the ending consonant as an intact unit. The performance of rhyme detection improved at time three for all students. There are two explanations. The first is that students may develop the insight of the onset-rhyme structure of English words through the growth of vocabulary (Metsala & Walley, 1998). The second is that explicit instruction on sound-print mapping may direct students to notice the consistent sound-print correspondence at rhyme level.   114  Spelling. Fixed effects in the model of spelling are shown in  Table 20 . Both groups improved significantly in spelling from time two through time three, as indicated by the significant parameter of time 3 (p < .001). The two groups attained comparable results at the immediate posttest, indicated by the non-significant parameter of group (p = .11), but were different at the delayed posttest (p = .01), with an effect size of 0.49. This indicated that the experimental group improved faster than the control group in spelling from time two through to time three.  The result is discussed in the following section, spelling error analysis. Table 20  Fixed effects in the model of invented spelling  Β SE df t p Intercept 11.19 2.03 6.91 5.50 < .001 Time 3 3.53 1.09 85.00 3.25 < .001 Exp group 2.49 1.69 5.23 1.47 .105 No prior EXLE -2.50 1.25 79.02 -2.01 .02 Exp group at t3 2.85 1.53 85.00 1.86 .01 Note. Exp, experimental; No prior EXLE, no prior experience of learning English.  Letter naming. Fixed effects in the model of letter naming is in Table 21. Letter naming improved significantly at time three (p < .001). Students who had English instruction before school entry had better performance at letter naming (p < .001). The instructional effects were not significant at both time points. Table 21  Fixed effects in the model of letter naming  Β SE df t p Intercept 7.43 1.92 11.44 3.86 < .001 Time 2 1.23 1.00 116.92 1.22 .11 Time 3 7.47 1.59 4.73 4.69 < .001 Exp group 0.70 1.71 8.89 0.41 .35 No prior EXLE -5.37 1.46 82.37 -3.68 < .001 Exp group at time 2 1.40 1.40 167.46 1.00 .16 Exp group at time 3 1.10 1.40 167.54 0.78 .22 Note. Exp, experimental; No prior EXLE, no prior experience of learning English.   115  This result is consistent with the previous studies of phonemic awareness training. Explicit instruction of phonological skills did not promote students’ letter knowledge (Yeung et al., 2013). Letter naming, a precursor to English literacy, is more related to students’ exposure to English print, and the home literacy environment (Foy & Mann, 2003).   Oral cloze. Fixed effects in the model of oral cloze is in Table 22. All students did not improve significantly in oral cloze until the delayed posttest (p < .001). Students who had English instruction before school entry performed significantly better on oral cloze. PA instruction was not effective on oral cloze at either time points. Table 22  Fixed effects in the model of oral cloze  Β SE df t p Intercept 2.62 0.42 21.09 6.26 <.001 Time 2 0.05 0.30 170.00 0.15 .44 Time 3 2.67 0.30 170.00 8.81 <.001 Exp group -0.01 0.42 7.73 -0.03 .49 No prior EXLE -0.61 0.33 80.26 -1.84 .04 Exp group at t2 0.56 0.43 170.00 1.78 .34 Exp group at t3 0.42 0.43 170.00 0.98 .17 Note. Exp, experimental; No prior EXLE, no prior experience of learning English.  Vocabulary. Fixed effects in the model of English picture naming is in Table 23. The extracurricular parameter was significant (p < .001), indicating that students who had previously learned English had a significant advantage on vocabulary. Students showed improvement in vocabulary in both post (p = .02) and delayed posttest (p < .001). The instructional effects were not significant.      116  Table 23  Fixed effects in the model of picture naming  Β SE df t p Intercept 5.18 0.63 17.80 8.23 <.001 Time 2 1.50 0.55 6.25 2.71 .02 Time 3 6.68 0.46 10.18 14.56 <.001 Exp group 0.73 0.57 32.09 1.28 .11 No prior EXLE -2.38 0.50 82.70 -4.73 <.001 Exp group at t2 0.18 0.60 165.01 0.30 .38 Exp group at t3 -1.10 0.60 165.16 -1.84 .06 Note. Exp, experimental; No prior EXLE, no prior experience of learning English.  Picture naming and oral cloze, both indicators of oral proficiency, did not improve under the influence of the PA instruction. This provides evidence for discriminative validity of the instruction. Phonological awareness and oral language proficiency are correlated but distinct constructs. Explicit instruction on one does not have a direct effect on the other. The pattern was verified in the previous study (Bowyer-Crane et al., 2008). Yeung (2013) investigated an intergraded instruction incorporating both oral language proficiency and PA skills. The instruction was found to have positive effects on expressive vocabulary. The effect size, however, was admittedly small.  Pinyin spelling. As shown in Table 24, all students improved significantly at time two, and both groups did not show any difference at either time points. Table 24  Fixed effects in the model of Pinyin spelling   Β SE t p Intercept 17.27 0.63 25.17 <.001 Time 2 1.53 0.59 2.87 .004 Exp group 0.39 0.69 0.56 .59 No prior EXLE -.60 0.54 -1.01 .27 Exp group at t2 0.37 0.73 0.54 .59 Note. Exp, experimental; No prior EXLE, no prior experience of learning English.   117  The distribution of the Pinyin spelling score was negatively skewed.  Almost of half of students (47%) attained full marks on Pinyin spelling at pretest. This was attributed to students’ previous experience of receiving intensive Pinyin instruction. According to the school principal, most of the students would have had Pinyin instruction in Kindergarten. The school heavily emphasized Pinyin instruction throughout the first six weeks of the school year, as they believed it was the foundational skill of Chinese literacy. However, there were students who fell behind in this skill, as 25% students attained less than 70% on the Pinyin spelling test.  A secondary analysis was conducted on the instructional effect among those who fell behind on Pinyin spelling. The results of the participants who performed below the 50th percentile rank in invented Pinyin spelling at pretest (Mean = 14.93, SD = 3.39) were selected, and an ANCOVA analysis was performed, with group being the independent variable, and the performance at pretest being the covariate. The results indicated that the group effect was significant (p = .03), with a small effect size (Cohen’s d = 0.27). Pinyin spelling of the two groups at time one and time two is presented in Figure 9. I speculated that the positive effect was due to students transferring the improvement in English phonemic awareness to Pinyin spelling. PA instruction gave these students opportunities to practice segmenting and blending phonemes, the skills of which were also involved in Pinyin spelling. A further experimental investigation of the causal relation is needed.  118   Figure 9 Pinyin spelling of students who fell behind the 50th percentile rank  Chinese character reading. As shown in Table 25, all students improved significantly at time two and at time three in Chinese character reading. Expectedly, the instruction of English phonological awareness did not have any effect on Chinese character reading at either time points. Table 25  Fixed effects in mode of Chinese character reading  Β SE t p Intercept 29.78 8.95 3.32 .001 Time 2 10.76 1.55 6.71 <.001 Time 3 27.32 1.70 16.61 <.001 Exp group -1.95 4.51 -0.48 .68 No prior EXLE 1.68 2.26 0.73 .47 Exp group at t2 0.98 2.32 0.41 .68 Exp group at t3 0.58 2.39 0.24 .81 Note. Exp, experimental; No prior EXLE, no prior experience of learning English.   119  Whether PA Instruction is a Necessary Prerequisite for Learning Letter-Sound Knowledge Letter-sound knowledge was the primary outcome of the phonics program in term two. Whether the PA instruction in term one had a delayed effect on letter-sound knowledge was investigated. A linear mixed model was employed for the analysis. Group, prior English learning experience, and letter naming at time one were set as fixed effects. Random intercept and random slope of group within teachers were set as random effects. The estimated fixed effects appear in Table 26. The results indicated that the experimental group performed slightly better than the control group on letter-sound knowledge at time three. However, the group effect did not reach statistical significance after students’ previous performance of letter naming and extra-curricular experience was controlled (p = .30).  Table 26  Fixed effects in models predicting letter sound knowledge at delayed posttest  Β SE t p Intercept 13.05 1.96 6.66 <.001 Ln t1 0.39 0.09 4.32 <.001 Exp group 2.39 2.20 1.13 .30 No prior EXLE -2.61 1.26 -2.02 .04 Note. No prior EXLE, no prior English instruction; Ln, letter naming; Exp, experimental.  Students in the experimental group received PA instruction between the preceding Pinyin instruction, and the following phonics instruction.  Pinyin and phonics instruction focused on systematic learning of alphabetic systems, however neither included the component of explicit instruction on phonological awareness. In terms of curriculum design, PA instruction was the missing piece to the puzzle. However, there are studies suggesting that systematic and explicit instruction on alphabetic principles could result in the development of PA (See Ziegler & Goswami, 2005 for a review). The present study results indicated the PA instruction gave students a head start on pseudoword reading, as well as phonemic awareness, however the advantages were negated  120  following systematic instruction of English alphabetic principles. In addition, phoneme awareness, the major outcome of PA instruction, did not predict students’ later knowledge of letter sounds, which was the primary learning outcome of phonics instruction. All these findings suggest that PA instruction may not be a necessary prerequisite for learning English phonics.  Therefore, instead of delivering PA training prior to a phonics program, which focuses on alphabetic principles, embedding PA activities in phonics instruction might be as effective, and more efficient. However, there is evidence suggesting that further explicit instruction on PA was necessary for children who were at risk of reading disabilities (Hatcher et al., 2006). Further studies on this topic are needed, for students who are at risk of failure in learning English as a foreign language. Spelling Error Analysis The average score of spelling response for each word is shown in Figure 10.  On average, students were able to spell the initial consonant at time two, and to spell two or more letters of a simple word at time three, following one term of instruction. I micro-coded every spelling response for the following elements; initial consonants, 2nd consonant, final consonant, medial vowels, and consonant clusters. The percentage of correct responses under each category is shown in Figure 11.  Position effect. A significant position effect was found (F (2, 330) = 206.12, p < .001). Using a Bonferroni-based procedure, with adjusted p =.01, a multiple comparison test was conducted. The student participants performed significantly better at spelling initial consonants (M =.73, SD =.29) than final consonants (M =.49, SD =.32), and they attained greater performance at spelling final consonants than medial vowels (M =.22, SD =.25). The pattern of position effect in spelling, which is that initial consonant > final consonant > medial vowel,  is  121  consistent with the position effect found in native English speaking children (Lee & Al Otaiba, 2017). In contrast, Hong and Chen (2011) found a different patterned position effect in Taiwanese children, which is that of initial consonant > medial vowel > final consonant. They argued that Chinese phonological structure does not allow final consonants, which might lead to poor phonemic representation of the final consonant in English. The pattern was only found with low proficiency students, who did not receive any explicit instruction on phonics. The participants in this present study have received instruction on alphabetic principles and Pinyin instruction. The experience of  learning alphabetic principles can enhance the presentation of phonemes at the ending position, both aurally and orthographically (Wang & Geva, 2003).   122   Figure 10 Average spelling response of each item of two groups at two time points 123   Figure 11 Percentage of correct responses on each element at posttest and delayed posttest 124  Group effect. The two groups were compared on phoneme spelling on three categories; single consonant, medial vowels, and consonant clusters. The percentage of correct responses under each category at both time points is presented in Figure 12. A two-way MANOVA was conducted, with spelling on single consonants, spelling on medial vowels, and spelling on consonant clusters being the dependent variables, group and time being the independent variables. The results indicated that the group and time effects were significant (p < .001 for both effects). To further investigate the group differences on each element, a post hoc analysis was conducted.  Three lineal mixed models were built for single consonant spelling, vowel spelling, and consonant cluster spelling respectively, with time and group being the fixed effects and student being the random effect.   Figure 12 Percentage of correct responses on each element of two groups at two time points  The models of spelling of the three elements appear in Table 27. The estimated model of single consonant spelling showed that students from both groups improved significantly from  125  time two to time three, and that the PA instruction had a significant effect, favoring the experimental group. The estimated model of vowel spelling indicated that the group had no effect at time two, however had a significant effect at time three. Students in the experimental group improved in vowel spelling, however students in the control group did not. The estimated model of consonant cluster spelling revealed the same pattern. The two groups were comparable at time two, however the experimental group performed significantly better at time three.    Table 27 Fixed effects in the models of spelling on the three elements  Model of SC Model of V Model of CC Intercept 0.96** 0.35** 0.66** Time 3 1.11** 0.05 0.14 Exp Group  1.01** 0.21 0.08 Exp Group at t3 0.17 0.38* 0.50* Note. SC single consonant, V vowel, CC consonant clusters; *, <.05; ** <.001.  It was not feasible to assess spelling at time one, as most students had no prior experience of the English language at this time. Spelling was not required and not explicitly taught until grade three, and 90% of students did not spell any of the words correctly at time two. However, almost all of the students could spell a part of an English word, showing their insights of the intra-structure of English words at time two. On average, the students were able to spell the initial letter of a word correctly. The spelling of consonant clusters and vowels was a difficult task for students of both groups. Although the two groups of students did not show differences in the level of sophistication in English spelling at time two, at the micro level students from the experimental group did present with an advantage on spelling single consonants. At time three, after students had received phonics instruction, the spelling performance of both groups improved, although students in the experimental group improved more than those in the control group. On average,  126  they were able to represent all of the phonemes in short words, of simple structure and utilizing conventional letters, although not always spelled correctly. This delayed instructional effect of PA instruction on spelling was found with native English speaking students (Tangel & Blachman, 1995).  The results of a spelling error analysis showed that the group effect at time three may be attributed to the students’ advantage in spelling vowels and consonant clusters, as both elements are particularly difficult to spell at the initial stages of spelling development (Treiman, 1991). Through receiving instruction on segmenting speech, and identifying and discriminating phonemes, the ability of phonologically encoding the challenging elements in spelling was enhanced. Invented spelling at time three moderately correlated with initial phoneme deletion at time two (r = 0.37).  Moderating Effects of Prior English and Chinese Variables The immediate effects of PA instruction were significant on English phonemic awareness, as well as pseudoword reading. The moderating effects of prior skills on the instructional effects were explored. Correlation coefficients between predictors at pretest and outcome variables at posttest of both experimental and control groups appear in Table 28. The strength of the association between some of the predictors and of the outcomes was shown to be largely different across the two groups. Picture naming, word reading and Pinyin spelling were shown to be more correlated with the outcomes in the experimental group.       127  Table 28  Partial correlation coefficients between predictors and outcomes after age was controlled   Pypre Chrpre Ocpre Vocpre widpre Experimental             Pswidpost 0.36* 0.32* .34* 0.38* 0.59* Intdepost 0.26 0.20 .09 0.38* 0.39* Control          Pswidpost 0.10 0.25 .18 -0.01 0.30 Intdepost 0.24 0.12 .02 -0.29 -0.09 Note. Pswidpost, pseudoword reading posttest; Intdepost, initial phoneme deletion posttest; Pypre, Pinyin spelling pretest; chrpre, Character identification pretest; ocpre, oral cloze pretest; vocpre, vocabulary pretest; widpre, word identifcation pretest. *, p <.05   The moderating effects of Pinyin spelling, picture naming, and English word reading were investigated. Since the distribution of Pinyin spelling was negatively skewed, a categorical variable was created for Pinyin spelling as described above.  Linear mixed-effects models were built for phonemic awareness, and pseudoword reading at immediate posttest respectively. Group, outcome performance at pretest, predictor performance at pretest, and interaction between group and predictor were set as fixed effects. Class was set as a random effect. As the analysis of the moderating effects was post hoc, correction of the alpha inflation was applied (Hochberg & Benjamini, 1990). The fixed effects in models of phonemic awareness appear in Table 29.The results indicated that the instructional effect on phonemic awareness was significantly moderated by English expressive vocabulary (Β =0.36, p = .01), as students’ prior performance on picture naming significantly predicted their later performance of initial phoneme deletion in the experimental group after controlling for students’ prior performance on initial phoneme deletion. In contrast, neither students’ prior Pinyin spelling skill nor English word recognition skill moderated the instructional effectiveness on phonemic awareness, as indicated by the non-significant interaction effects.  128  Table 29  Fixed effects in models of phonemic awareness with Pinyin spelling, picture naming and word recognition as moderator  Β SE t p Model1     (Intercept) 0.51 0.32 1.59 .12 Expgroup 1.20 0.54 2.21 .03 Pypre 0.12 0.56 0.22 .83 Intdepre 0.83 0.12 7.18 < .001 expgroup1:pypre -0.10 0.78 -0.12 .90 Model 2     (Intercept) 1.03 0.45 2.27 .03 expgroup1 -0.13 0.61 -0.21 .83 Vocpre -0.15 0.12 -1.19 .24 Intdepre 0.76 0.11 6.98 < .001 expgroup1:vocpre 0.36 0.14 2.51 .01 Model3     (Intercept) 0.63 0.37 1.70 .09 expgroup1 1.02 0.51 1.98 .05 Widpre -0.02 0.07 -0.27 .79 Intdepre 0.80 0.13 6.05 < .001 expgroup:widpre 0.04 0.09 0.46 .64 Note. Expgroup, experimental group; Intdepre, initial phoneme deletion pretest; Pypre, Pinyin spelling pretest; vocpre, vocabulary pretest; widpre, word identifcation pretest.  The fixed effects in models of pseudo word reading appear in Table 30.The instructional effect on pseudoword reading was significantly moderated by Pinyin spelling (Β= 2.80, p =. 04), as students’ prior performance on Pinyin spelling significantly predicted their later performance of pseudoword reading in the experimental group. Neither prior English expressive vocabulary nor English word recognition moderated the instructional effectiveness on English pseudoword reading.       129  Table 30   Fixed effects in models of pseudoword reading with Pinyin spelling, picture naming and word recognition as moderator  Β SE t p Model1     (Intercept) 1.36 1.02 1.33 .21 expgroup1 0.72 1.48 0.49 .63 Pypre 0.90 1.16 0.78 .44 Pswidpre 0.54 0.10 5.71 < .001 expgroup1:pypre 2.80 1.60 1.75 .04 Model 2     (Intercept) 2.07 1.16 1.79 .08 expgroup1 1.65 1.55 1.07 .30 Vocpre -0.15 0.28 -0.53 .60 Pswidpre 0.56 0.12 4.84 < .001 expgroup1:vocpre 0.28 0.32 0.88 .38 Model3     (Intercept) 1.63 0.93 1.76 .10 expgroup1 1.67 1.27 1.32 .11 Widpre 0.17 0.15 1.14 .26 Pswidpre 0.35 0.13 2.81 .01 expgroup1:widpre 0.17 0.17 1.04 .20 Note. Expgroup, experimental group; Intdepre, initial phoneme deletion pretest; Pypre, Pinyin spelling pretest; vocpre, vocabulary pretest; widpre, word identifcation pretest.  The significant moderating effect of picture naming on initial phoneme deletion suggests possible causal influence of oral language ability on phonemic awareness. This finding supports the lexical restructuring model (Metsala & Walley, 1998), which proposed that development of explicit phonological awareness derives from the growth of vocabulary.  The moderating effect of Pinyin spelling on the instructional effect on pseudoword reading provided evidence for cross-linguistic transfer of decoding skills. Students with greater L1 literacy abilities were more successful at transferring the English phonemic awareness towards decoding pseudowords. The moderating effect was absent on the instructional effect on phonemic awareness, which was the main focus of the instruction, indicating that students’ gain of English phonemic awareness was not related to their prior Chinese decoding abilities.   130  This study used experimental manipulation to investigate cross-linguistic transfer. The findings contradicted Goodrich, Lonigan, and Farver (2012)’s study, where the print knowledge of Spanish, including letter-sound-corresponding rules, and word reading, did not moderate the instructional effect on English print knowledge. One explanation of the contradiction could be the difference of proficiency levels of students between the two studies. According to the threshold hypothesis, transfer of language and literacy skills only occurs when students have reached certain a level of proficiency in at least one language. Participants in the Spanish study were enrolled in kindergarten, and whose literacy proficiency for both languages were low. Participants in the present study were already established in Chinese oral language, and had started intensive instruction of Chinese literacy. Thus they were able to transfer their Chinese literacy skills to English literacy outcomes, as suggested by the association among Chinese Pinyin and character identification skills, and English word reading. Chinese script is distinct from English script, whereas Pinyin script much resembles English script. The ability in Pinyin spelling shows relatively stronger association with English pseudoword reading.   Findings in Goodrich et al. (2012)’s study suggested that phonological awareness in Spanish predicted students’ benefits derived from the literacy instruction on English phonemic awareness among ESL students whose first language was Spanish. The instructional effect on English phonemic awareness was moderated by their prior Spanish phonemic awareness. In the present study, students’ prior phonological awareness of Chinese was not assessed. If done so, it might moderate the effect on phonological awareness.  Limitations The present study had the following limitations. Firstly, I failed to include a comparison group to whom neither PA nor phonics instruction was delivered. Thus it cannot be known  131  whether the instructional effects would remain at the delayed posttest, if no phonics instruction was introduced. The reason that the control group’s performance reached that of the experimental group at time three could be due to the maturity, or natural exposure to English, and not the phonics instruction.  Secondly, phoneme blending, one of the instructional outcomes, was not included as part of the assessment. It is an important skill for learning synthetic phonics, which requires students to identify the sound of each letter, and blend the sounds to create a word. It is hard to predict the instructional effect on phoneme blending based on performance of the other tasks, as the results showed that students’ performance of phonemic awareness varied greatly across tasks.  Implications for Practice Phonological awareness should be included in English instruction to young students in China who are taught English as a foreign language. Auditory analysis of speech sounds without using letters has positive effects on learning to read and spell in English. The greatest strategy would be to deliver the instruction in small doses and frequently, and can certainly be delivered in various game formats to engage young students. Among the several instructional components, phoneme segmentation and blending were found to be the most effective with Chinese children. The PA instruction creates opportunities for children to practice analyzing speech sounds, thus they become aware of sound structures in English. This further enhances the skill of English spelling, particularly the spelling of complex phonological elements such as consonant clusters and vowels. Knowledge and skills in Pinyin can help students learn English alphabetic principles, word reading and spelling. Teaching phonics after students have mastered the skill in Pinyin has shown to be effective in this study.   132   Letter-sound knowledge and phonemic awareness, both elements important to English reading and spelling, can be taught concurrently, as acquiring one does not depend on the knowledge of the other. Embedding the PA components in a phonics program may be the best practice for young Chinese EFL students regarding instructional efficiency. Conclusion PA instruction was shown to be effective on skills including phonemic awareness, pseudoword reading, and spelling, however not effective on rhyme detection, letter knowledge, word reading, nor oral language skills. The instructional outcome of phonemic awareness was moderated by prior expressive vocabulary. Students who had greater expressive vocabulary were shown to have greater gains of phonemic awareness. The instructional effect on pseudoword reading was moderated by Chinese literacy outcomes, particularly by Pinyin spelling. Students who were good Pinyin spellers in pretest were able to transfer the outcome of PA training to novel word decoding. PA training was not a necessary precursor to learning letter-sound knowledge for two reasons; first, students who did not receive prior PA instruction caught up in word reading at the delayed posttest, and second, phonemic awareness prior to the instruction of letter-sound knowledge did not predict subsequent instructional outcomes.     133  CHAPTER 5: GENERAL DISCUSSION AND CONCLUSIONS The study series was designed with the purpose of providing evidence to inform educators of early English language learning for young learners in China, and other non-English speaking foreign countries, particularly in the instruction of reading. The study was also designed to answer theoretical questions regarding reading and bilingualism, such as whether phonemic awareness can have a causal influence on word reading among EFL learners, whether the language learning environment influences development of foreign language skills, and whether literacy skills in students’ native language can predict a student’s ability in learning to read in English. Development of English Phonological Awareness among Chinese EFL Children Phonological awareness were measured by two types of tasks in present research, detection and deletion. Detection taps into awareness of implicit segmental structure of words (Walley et al., 2003). Deletion measures the ability of explicit manipulation of phonemes. Both studies showed that the children’s performance of initial phoneme detection reached a ceiling effect at the initial data collection time, time point one, before the delivery of the phonological awareness instruction. The ability of explicitly manipulating English phonemes, on the other hand, takes longer to develop.  Both studies showed that the children’s performance of initial phoneme deletion almost reached the floor effect at the onset of the study. These findings confirmed with the lexical restructuring model that the implicit and perceptual concept of phoneme is a precursor to the explicit segmentation or phoneme awareness skills (Metsala & Walley, 1998; Walley et al., 2003).  Both studies showed that expressive vocabulary predicted the development of explicit phonemic awareness. Study 1 indicated that children with more expressive vocabulary attained  134  greater performance of explicit phonemic awareness, when again measured three months after the start of the study. Study 2 showed that children with better expressive vocabulary attained greater gains from the phonological awareness instruction. These findings are consistent with the lexical restructuring model, which proposed that explicit phonemic awareness is derived from the growth of vocabulary in the target language (Metsala & Walley, 1998). Students attained near perfect performance on the detection of the initial sound of English words, however showed much poorer performance on detecting rhymes. The detection of rhyme units should be easier than the detection of the initial consonant, as the rhyme unit shares a larger global similarity with the target word than phonemes (Ziegler & Goswami, 2005).  However, both studies indicated the opposite pattern presented in the Chinese children who participated the study. The explicit Instruction of rhyme awareness in Study 2 was found not to be effective, a students from both groups did not improve their rhyme detection abilities when measured at time point two, in contrast to their significant improvement of vocabulary and word reading. A possible explanation could be that English vowels, which are the principal components of rhyme units, are not well represented among the Chinese language, as the Chinese language has fewer vowel categorizations (Howie, 1976). The same pattern was revealed in children’s spelling responses. Participants, from both immersion and non-immersion programs, were very proficient in spelling consonants, however presented with rudimentary performance and skills regarding the spelling of vowels. In addition, onset-rhyme segmentation of a syllable is not a natural process for Chinese speakers, even with experience in learning English (Chen, 2011),  as the body-coda segmentation of a syllable, rather than onset-rhyme is more natural for Chinese speakers. Therefore, their performance of rhyme detection was poor.  135  Influence of Phonological Awareness on English Reading among Chinese Speaking Children The two studies revealed a converging pattern that earlier performance of initial phoneme deletion was associated with English word reading. However, the causal influence of phonemic awareness on English word reading at an early stage is not established in either study. The cross-lagged correlation analysis conducted as a part of study 1 showed that performance of word reading at time 1 was more associated with the later performance of phonemic awareness at time 2 than the reverse order, indicating that the prior phonemic awareness did not affect the subsequent performance of word reading. In addition, the instruction on phonemic awareness in Study 2 had no immediate or delayed effect on real word reading.  These findings contradict the studies with English L1 children, which showed that phonemic awareness is a very important precursor of literacy (e.g. Firth,1985) and has a causal influence on word level reading (e.g. Hulme et al., 2012). EFL students in this study, similar to most other EFL students, lacked the language exposure and oral proficiency to develop English phonemic awareness before they learned to read and write English. Thus it is impossible that English phonemic awareness serves as the precursor to literacy. The present research, together with Yeong and Rickard Liow (2011) ‘s study, suggests that it is the experience of learning English literacy that causes the development of English phonemic awareness, not the reverse.  Despite the findings indicating phonemic awareness did not directly cause an improvement in word reading, it is an important underlying skill for word recognition. Instruction of phonemic awareness had a positive effect on pseudoword reading, indicating a direct influence on phonological decoding which is an essential skill for word recognition and reading (Hoover & Gough, 1990). English phonemic awareness, once established, may serve as  136  an underlying skill for reading among students in subsequent school grades. The results of the cross-lagged analysis performed as part of study 1 showed that phonemic awareness at time 2, which was more refined than at time 1, had a significant effect on word reading at time3. This finding is consistent with the pattern found by Hu and Schuele (2015) that EFL students in grade 5 who performed below expectations on English word reading were present with a mild deficit in English phonological awareness assessed at grade 4. Both studies showed that earlier phonemic awareness has significant influence on later attainment of word reading.       Implicit phonemic awareness was shown not to be associated with English word recognition. However, this finding does not indicate if implicit phonemic awareness is irrelevant to English word reading for Chinese-speaking children. Performance of initial phoneme detection showed a ceiling effect in this study, indicating that most of the participants had sufficiently mastered this skill. The predictive power of this skill regarding word reading may be revealed among younger students. A study with Chinese speaking kindergartner participants in Hong Kong showed that the task of phoneme identification was the most powerful predictor to later English word identification (Yeung & Chan, 2013). The association of rhyme awareness and word reading, however, did not show a consistent pattern across the two studies. Study 1 showed that rhyme awareness at time 1 was associated with later performance of phonemic awareness and word reading.  This finding is consistent with Yeung’s (2010) study, which indicated that rhyme detection had a direct influence on word reading. However, the pattern of association of rhyme awareness with word reading skills was not present in the grade 1 sample in study 2. Rhyme awareness at pre-test was not associated with any English literacy measures at any study time point.   137  Study 1 found that phonological awareness was only shown to be associated with word decoding, and was only tangentially associated with semantic access. The dissociation of these two subcomponents of word recognition in terms of underlying predictors supports the statement that English word decoding and meaning access are two loosely connected constructs among beginning EFL learners (Jiang, 2000; Saiz 2007). Most of the previous studies on EFL word recognition focused on word decoding. Findings in this study add to the literature which investigated the association between phonological awareness and access of the meaning of words (Allen-Tamai, 2000). However, the previous findings are in contradiction with the results ascertained in study 1. More studies are needed in the future to draw a conclusion. Effectiveness of Phonological-Based Instruction among EFL Students The present research investigated the effectiveness of two types of phonological-based instruction, phonological awareness instruction and phonics instruction, among Chinese young EFL learners. Both served the purpose of cultivating the skill of phonological decoding of English words. The findings suggest that EFL students in lower grades can benefit from phonological-based instruction to acquire insights into English phonemic structure and alphabetic principles, despite their limited language exposure and oral language proficiency. Phonics instruction was provided for grade 2/3 students in Study 1, in addition to the current curricula.  Although no comparison condition existed as a control measure for the phonics instruction, there was evidence that suggested at the effectiveness of the phonics instruction. It was found that the performance of pseudoword reading of student participants of both programs increased significantly after the receiving phonics instruction. This improvement can be attributed the phonics instruction, at least in part, as the prior English curricula in both  138  programs did not contain explicit instruction of phonemic awareness or phonetics, and that the performance of pseudoword reading was shown to be negligible before receiving instruction.  Phonological awareness instruction, as part of study 2 was provided for grade 1 students as a prerequisite to phonics instruction. The phonological awareness training in study 2 was compared to regular pre-existing morning reading sessions, which consisted of components such as oral reading, repeated reading, and whole word recognition. It was found that small dosages of phonological awareness training did not develop an advantage for real word reading, compared to the status quo, however the instruction was shown to have a moderate effect on phonemic awareness and a moderate transfer effect on pseudoword reading.   Several factors were identified that influenced effectiveness of the instruction. Enriched language learning context facilitates learning phonics. Study 1 showed that the advanced level of phonics instruction, which focused on complex sound-print-corresponding patterns, was more effective with students participating in the immersion program. It was speculated that students in the immersion program were more likely to be exposed to words with complex spelling patterns in their daily reading, and thus were more prepared to receive the explicit instruction of the spelling rules, and in turn having greater opportunities to practice decoding complex words. English oral language proficiency, specifically expressive vocabulary influenced students’ gain from the phonological awareness instruction.    Another influencing factor is the pre-existing Pinyin skill. Students with greater Pinyin skills were better able to transfer the outcome of English phonemic awareness to the act of reading pseudowords. Many studies have suggested that literacy skills of one’s first language, such as decoding and phonological awareness, predict the development of literacy skills of a second language (e.g. Shum, Ho, Siegel, & Au, 2016). This is the first study which adopted  139  experimental manipulation to investigate the effect of decoding skills of the first language on the learning of decoding skills in a foreign language.  Study 2 also suggested a sequence of delivering the instructional components. Phonological awareness training, which is presumed to be a precursor to phonics instruction (Groff, 2001),  however, was shown in this study to have no sustained effect on phonological awareness or pseudoword reading five months following the initial instruction. The control group, who only received instruction on alphabetic principles without training on phonological awareness, caught up with most of the outcomes. This suggested that phonological awareness training does not necessarily antecede instruction of alphabetic principles.  However, the finding does not support the statement that PA training is redundant, and does not add value to the overall effectiveness if alphabetic principles are explicitly taught. PA activities were embedded in the alphabetic principles instruction as part of Study 2, as reported by the teacher participants. The improvement students made reflect the efficacy of these two components, and not only alphabetic principles. Moreover, spelling error analysis showed that students who received extra PA training attained greater performance at spelling phonemes that were more challenging to represent phonologically, such as vowels and consonant clusters. These findings suggest a retained effect of PA training on spelling. This finding is consistent with  previous findings among English native speaking children that phonemic awareness is more related to spelling performance than word reading (Ellis, 1997). The instruction in this study was delivered at the whole-class level, with large group sizes. The  phonological-based instruction received teacher-centric adaptation, entailing skill drill exercises. Use of Powerpoints, pictures, and motor activities was frequently adopted to promote student engagement throughout the instruction. Previous intervention studies, which indicated  140  large effects, were conducted with students who were struggling readers, in either group or individual settings (Ehri, Nunes, Stahl, et al., 2001). The findings of this study add to the evidence that phonological based instruction, in an EFL context, is effective on phonological decoding skills at a whole-class level (Dixon, Schagen, & Seedhouse, 2011; Lai et al., 2009; Yeung et al., 2013).  Cross-Linguistic Transfer  Both studies found evidence of cross-linguistic transfer of decoding abilities. Study 1 indicated that Chinese character identification predicted word pronunciation and semantic access across a year. Study 2 found that students who were superior performers of Pinyin spelling at pretest made greater gains in pseudoword decoding after having received phonological awareness training. Cross-linguistic transfer of decoding skills in this study could be enabled by the environment that supports learning of both languages. Students in the immersion program spent no less time than students in the non-immersion program on learning Chinese literacy, which qualified the program as an additive bilingual program. Another factor that may enable the transfer of decoding skills is the process of phonological analysis, which underlies the decoding of words in both English as well as Chinese Pinyin. The process of phonologically analyzing the word, and applying the rules of print-sound conversion to sound out the word, is a process shared in both Pinyin decoding and English pseudoword reading. Study 2 illuminated the underlying mechanism of cross-linguistic transfer of decoding abilities, as students who received phonological awareness training showed a stronger association between Pinyin spelling and English psuedoword reading. This result suggests that phonological training promotes English word decoding through activating the  141  transfer of knowledge and skill of deciphering the alphabetic system in the students’ first language. The phonological awareness training had less influence on the transfer of Chinese character identification ability to pseudoword reading, possibly as the phonological analysis was less dominant in Chinese character identification. The covariance between Chinese character identification and English reading needs to be further addressed in future studies. The instruction in study 2 did not show an effect of reverse transfer, from English to Chinese, mainly as most of the students had shown a ceiling effect for Pinyin spelling. By examining students who were less proficient in Pinyin spelling at pretest, it was found that their Pinyin spelling did slightly improve in response to English phonological awareness instruction, compared to students in the control condition. This suggests that training of English speech analysis has a positive effect on the same ability in Chinese. Chinese explicit phonological awareness was not assessed in this study. We speculated that the training would have a positive effect on explicit Chinese awareness too.    The English immersion program, as it is designed at present, treats the learning of the two languages separately, and therefore the use of Chinese is prohibited while in the English class. This pedagogical methodology is in part due to the two solitude assumption, as postulated in the direct method of teaching foreign languages (Richards & Rodgers, 2014).  Although not encouraged in teaching, the findings showed that decoding skills are strongly correlated across languages, particularly for students enrolled in the immersion program. As advocated by Cummins (2005) , efforts should be made to investigate the efficacy/effectiveness of programs that are teaching for transfer. Cummin made suggestions in regards to the incorporation of different languages in one class, such as the development of dual language multimedia books,  142  words comparison, and using translation as a tool to compare and contrast across languages (Cummins, Brown, & Sayers, 2007).  Implications for Practice  Our findings have some implications for teaching English reading to young children in EFL context. The first one is that the roles phonological awareness and phonological decoding play in learning to read in English should be acknowledged in the field of English education for young EFL students. Phonological-based instruction is effective in helping EFL learners to gain linguistic insights of English speech and phonologically decode English words, the skills of which predict subsequent word decoding. Therefore, allocating resources to this instruction is necessary.  In the meantime, the phonological-based instruction may have limited direct effects in real word reading and meaning access. Therefore, it is likely that phonological decoding and word meaning need to be taught (at least to some extent) independently. One cannot assume that if the phonological processing skills are taught, that meaning access will naturally follow. Instructional components targeting word meaning and real word identification should be included in word level instruction.  The second implication is that Pinyin instruction can serve as an effective prerequisite of learning English phonological processing skills. A positive transfer effect from Pinyin spelling to English pseudoword reading was shown in this study. This finding suggests that the experience of learning Chinese Pinyin can enhance students’ ability to learn the written system of English, rather than causing a confusing barrier.  The third implication is that phonological-based instruction should be accompanied by oral vocabulary instruction to maximize the instructional effectiveness, as EFL students with  143  better oral English proficiency benefited more from the phonological awareness instruction in improving English phonemic awareness. Teachers and education administrators should recognize that oral language abilities, particularly expressive vocabualry, are the foundations of learning to read and developing phonemic awareness in EFL children. The development of oral language abilities should be the primary goal of early English education. This leads to the last implication that EFL students’ language learning context needs to be taken into consideration in teaching foundational reading skills. The effectiveness of explicit phonological-based instruction will be constrained if reading and other language activities are not supported outside of the phonological-based program.  Implications for Future Research The two studies conducted as part of this research provided evidence for the effectiveness of phonological-based instruction. Generalization of the findings, however, is limited due to sampling bias, as the current research used convenience sampling, and only one school was involved in each study. The effectiveness of the instruction should be evaluated on a far larger scale, involving a larger quantity of teacher participants, classes, and schools. The results of such a study should be powerful enough to investigate the effects of factors that potentially influence the instructional effectiveness, such as teacher proficiency, school SES and instructional setting. Future studies should systematically compare different conditions of phonological based instruction, such as implicit versus explicit instruction, and extra PA instruction versus embedded PA instruction in the EFL context. The PA instructional effectiveness should be investigated against different comparison conditions, as PA was used for various purposes, such as learning new vocabulary, and correcting pronunciation. The efficacy and effectiveness of the approach on vocabulary learning, specifically on print-meaning association and phonology- 144  meaning association, should also be investigated. The phonological-based approach should be compared against the commonly adopted vocabulary teaching approaches such as mnemonics and first language translation.  In order to further develop educational practices, future research should explore the possibility of teaching for language transfer, as opposed to keeping the instruction of two languages separate in both process and outcomes. This research, together with many previous studies, indicates that language transfer occurs with students of low and high proficiency.  The experience of one language enhances the ability and proficiency of the other. Instruction driven by this principle should be explored and evaluated. Instructional components such as comparing and contrasting languages were suggested to be included in foreign language education. Nishanimut, Johnston, Joshi, Thomas, and Padakannaya (2013) explored teaching the English written system analogous to the written system of the students’ native language. Grade 5 students from India were taught English synthetic phonics using the Kannada akshara, the written system of their native language, as an aid. For example, for the letter ‘z’, students would hear the sound /z/ spoken out loud by the teacher, and then the equivalent would be represented by a Kannada akshara  as well as the English letter “z/Z”. The results showed that the bi-literacy approach was more effective than the English only approach, as utilized in the control group.  Similar to the incorporation of the Kannada akshara script into English Phonics instruction, Pinyin scripts could serve as an aid for teaching English alphabetic principles to Chinese speaking EFL students. Future studies could be conducted to evaluate such a program that incorporates Pinyin with the teaching of English alphabet.  145  Conclusions Learning to read and spell in English is challenging for children in China, as they receive limited English language input in their daily lives and have low English oral proficiency. Phonological-based instruction have been shown to be effective in improving word-level reading skills among beginning readers. There is an insufficiency of evidence to support the instructional effectiveness among EFL children. Two factors which could potentially influence the instructional effectiveness were considered in this research. One is the environmental factor which is the educational program students were enrolled in to learn the English language. The other is the student factor which is the preexisting abilities of Chinese literacy and English oral proficiency. Given all the findings considered, I concluded that English phonological processing skills including phonological awareness and phonological decoding, the instructional outcomes of phonological-based instruction, significantly predicted the students’ performance in reading real words concurrently and six months later, after controlling for prior English vocabulary. This prediction stands regardless of whichever program students were involved in. In the meantime, English oral vocabulary is also important to word reading, such that it not only predicted the students’ subsequent performance in word decoding but also meaning access. The English language learning environment did have an impact on students’ learning of phonological decoding. Students in the immersion program made significantly larger gains in decoding pseudowords with complex spelling patterns, under the influence of systematic instruction of phonics, while the students from the two programs improved equally in decoding pseudowords with simple spelling patterns.  146  In terms of instructional effectiveness, results in study 2 indicated that the phonological-based instruction had immediate positive effects on initial phoneme deletion and pseudoword reading, and had a delayed effect on English word spelling. Chinese Pinyin spelling at pretest significantly moderated the instructional effect on pseudoword reading. English expressive vocabulary at pretest significantly moderated the instructional effect on phonemic awareness.   These results collectively suggest that the phonological-based approach, which is shown to be effective in improving word reading abilities among children in English dominant societies, had positive effects in improving English phonological processing skills among young EFL children in China, but the instructional effect on real word reading was limited. The instructional effects were enhanced in students who had relatively good English oral vocabulary and learned English in an enriched language environment. The educational implications are 1) to include phonological-based instruction in early English language education to the end of preparing students to learn to read in English, 2) components of words’ meaning and oral vocabulary should be emphasized in English reading instruction, 3) instruction on Chinese Pinyin enhanced rather than impeded learning the English written system. Future research should focus on investigating the effectiveness of specific instructional components in phonological-based instruction, and exploring the possibility of integrative instruction which combines oral language abilities and phonological skills, as well as Chinese Pinyin and English phonics.  147  BIBLIOGRAPHY Aaron, P. G., Joshi, R. M., Ayotollah, M., Ellsberry, A., Henderson, J., & Lindsey, K. (1999). Decoding and sight-word naming : Are they independent components of word recognition skill ? Reading and Writing: An Interdisciplinary Journal, 11, 89–127. http://doi.org/10.1023/A:1008088618970 Adams, J. (1990). Beginning to read: Thinking and learning about print. Cambridge: MIT Press. Allen-Tamai, M. (2000). Phonological awareness and reading development of young Japanese learners of English. Retrieved from ProQuest Dissertations and Theses. Temple University, Ann Arbor. Angiulli, A. D., Siegel, L. S., & Maggi, S. (2004). Literacy instruction , SES , and word-reading achievement in English-language learners and children with English as a first language : A longitudinal study. Learning Disabilities Research & Practice, 19(4), 202–213. http://doi.org/10.1111/j.1540-5826.2004.00106.x Asher, J. J. (1969). The total physical response approach to second language learning. The Modern Language Journal, 53(1), 3–17. Bates, D., Maechler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 1–48. Bergeron, B. S. (1990). What does the term whole language mean? Constructing a definition from the literature. Journal of Reading Behavior, 22(4), 301-329. Bialystok, E., McBride-Chang, C., & Luk, G. (2005). Bilingualism, language proficiency, and learning to read in two writing systems. Journal of Educational Psychology, 97(4), 580–590. http://doi.org/10.1037/0022-0663.97.4.580 Bowman, M., & Treiman, R. (2002). Relating print and speech: The effects of letter names and word position on reading and spelling performance. Journal of Experimental Child Psychology, 82(4), 305–40. Bowyer-Crane, C., Snowling, M. J., Duff, F. J., Fieldsend, E., Carroll, J. M., Miles, J., … Hulme, C. (2008). Improving early language and literacy skills: Differential effects of an oral language versus a phonology with reading intervention. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 49(4), 422–32. http://doi.org/10.1111/j.1469-7610.2007.01849.x Cain, K. (2007). Syntactic awareness and reading ability: Is there any evidence for a special relationship? Applied Psycholinguistics, 28(4), 679–694. http://doi.org/10.1017/S0142716407070361 Carroll, J. (2004). Letter knowledge precipitates phoneme segmentation, but not phoneme invariance. Journal of Research in Reading, 27(3), 212–225. http://doi.org/10.1111/j.1467-9817.2004.00228.x Chen, A. S. W. (2011). Bodies and codas or core syllables plus appendices? Evidence for a developmental theory of subsyllabic division preference. Cognition, 121(3), 338–362. http://doi.org/10.1016/j.cognition.2011.08.006 Cheng, L., Li, M., Kirby, J. R., Qiang, H., & Wade-Woolley, L. (2010). English language immersion and students’ academic achievement in English, Chinese and mathematics.  148  Evaluation & Research in Education, 24(2), 163–163. http://doi.org/10.1080/09500790.2011.579400 Cheung, H., Chen, H., & Yip, C. (2001). The development of phonological awareness : Effects of spoken language experience and orthography. Cognition, 81, 227–241. Chi, Y.,  Lan, J., & Zhang, L. (Eds.). (2010). Easy to learn to read in English: Phonics series in the English immersion program. Xi'an, China: Shaanxi Normal University General Publishing House. Chow, B. W.Y., McBride-Chang, C., & Burgess, S. (2005). Phonological processing skills and early reading abilities in Hong Kong Chinese kindergarteners learning to read English as a second language. Journal of Educational Psychology, 97(1), 81–87. http://doi.org/10.1037/0022-0663.97.1.81 Coltheart, M., Rastle, K., Perry, C., Langdon, R., & Ziegler, J. (2001). DRC: A dual route cascaded model of visual word recognition and reading aloud. Psychological Review, 108(1), 204–56. Cummins, J. (1979). Linguistic Interdependence and the Educational Development of Bilingual Children. Review of Educational Research, 49(2), 222–251. Cummins, J. (1998). Immersion education for the millennium: What have we learned from 30 years of research on second language immersion? In M. R. Childs & R. M. Bostwick (Eds.) Learning through two languages: Research and practice. Second Katoh Gakuen International Symposium on Immersion and Bilingual Education. (pp. 34-47). Katoh Gakuen, Japan. Cummins, J. (2005). Teaching for cross-language transfer in dual language education: Possibilities and pitfalls. Tesol, 1–7. Cummins, J. (2008). Teaching for transfer: Challenging the two solitudes assumption in bilingual education. Encyclopedia of Language and Vducation, Volume 5: Bilingual Education, 5, 65–75. http://doi.org/10.1007/978-0-387-30424-3_116 Cunningham, A. E., & Stanovich, K. E. (1997). Early reading acquisition and its relation to reading experience and ability 10 years later. Developmental Psychology, 33(6), 934–945. http://doi.org/10.1037/0012-1649.33.6.934 De Groot, A., & Keijzer, R. (2000). What is hard to learn is easy to forget: The roles of word concreteness, cognate status, and word frequency in foreign‐language vocabulary learning and forgetting. Language Learning, 50(1), 1–56. Dixon, P., Schagen, I., & Seedhouse, P. (2011). The impact of an intervention on children’s reading and spelling ability in low-income schools in India. School Effectiveness and School Improvement, 22(4), 461–482. http://doi.org/10.1080/09243453.2011.625125 Ehri, L. C. (1987). Learning to read and spell words. Journal of Literacy Research, 19(1), 5–31. http://doi.org/10.1080/10862968709547585 Ehri, L. C. (1992). Grapheme-phoneme knowledge is essential for learning to read words in English. In Word Recognition In Beginning Literacy (pp. 223–260). Ehri, L. C. (2014). Orthographic mapping in the acquisition of sight word reading, spelling memory, and vocabulary learning. Scientific Studies of Reading, 18(1), 5–21.  149  http://doi.org/10.1080/10888438.2013.819356 Ehri, L. C., Nunes, S. R., Stahl, S. A., & Willows, D. M. (2001). Systematic phonics instruction helps students learn to read: Evidence from the National Reading Panel’s meta-analysis. Review of Educational Research, 71(3), 393–447. http://doi.org/doi:10.3102/00346543071003393 Ehri, L. C., Nunes, S. R., Willows, D. M., Schuster, B. V., Yaghoub-zadeh, Z., Shanahan, T., … Lesegremiums. (2001). Phonemic awareness instruction helps children learn to read : Evidence from the National Reading Panel ’ s meta-analysis. Reading Research Quartely, 36(3), 250–287. http://doi.org/10.3102/00346543071003393 Ellis, N. (1997). Interactions in the development of reading and spelling: Stages, strategies, and exchange of knowledge. In C. A. Perfetti, L. Rieben, & M. Fayol (Eds.), Learning to spell (pp. 271–294). Hillsdale, NJ: Erlbaum. Fielding-Barnsley, R. (1997). Explicit instruction in decoding benefits children high in phonemic awareness and alphabet knowledge. Scientific Studies of Reading, 1(1), 45–63. http://doi.org/10.1207/s1532799xssr0101 Figueredo, L. (2006). Using the known to chart the unknown: A review of first-language influence on the development of English-as-a-second-language spelling skill. Reading and Writing, 19(8), 873–905. http://doi.org/10.1007/s11145-006-9014-1 Foy, J. G., & Mann, V. (2006). Changes in letter sound knowledge are associated with development of phonological awareness in pre-school children. Journal of Research in Reading, 29(2), 143–161. http://doi.org/10.1111/j.1467-9817.2006.00279.x Foy, J., & Mann, V. (2003). Home literacy environment and phonological awareness skills in preschool children. Applied Psycholinguistics, 24, 59–88. http://doi.org/10.1017/S0142716403000043 Gadermann, A. M., Guhn, M., Zumbo, B. D., & Columbia, B. (2012). Estimating ordinal reliability for Likert-type and ordinal item response data : A conceptual , empirical , and practical guide. Ptactical Assessment, Research & Evaluation, 17(3). Ganschow, L., Sparks, R. L., & Javorsky, J. (1998). Foreign language learning difficulties: An historical perspective. Journal of Learning Disabilities, 31(3), 248–258. http://doi.org/10.1177/002221949803100304 Gersten, R., Chard, D. J., Jayanthi, M., Baker, S. K., Morphy, P., & Flojo, J. (2009). Mathematics instruction for students with learning disabilities: A meta-analysis of instructional components. Review of Educational Research, 79(3), 1202–1242. http://doi.org/10.3102/0034654309334431 Geva, E. (2006). Second-language oral proficiency and second-language literacy. In D. August & T. Shanahan (Eds.), Developing literacy in second-language learners: Report of the National Literacy Panel on Language—Minority children and youth (pp. 123–139). Mahwah, NJ: Erlbaum. Geva, E., Yaghoub-zadeh, Z., & Schuster, B. (2000). Understanding individual differences in word recognition skills of ESL children. Annals of Dyslexia, 50(2), 123–154. Goodman, K. S. (1967). Reading: A psycholinguistic guessing game. Literacy Research and Instruction, 6(4), 126-135.  150  Goodman, K. S. (1986). What's whole in whole language? A parent/teacher guide to children's learning. Portsmouth: Heinemann Educational Books, Inc. Goodrich, J. M., Lonigan, C. J., & Farver, J. M. (2013). Do early literacy skills in children's first language promote development of skills in their second language? An experimental evaluation of transfer. Journal of Educational Psychology, 105(2), 414-430.http://doi.org/10.1016/j.pestbp.2011.02.012. Goswami, U. (1991). Learning about Spelling Sequences: The Role of Onsets and Rimes in Analogies in Reading. Child Development, 62(5), 1110. http://doi.org/10.2307/1131156 Gottardo, A., Yan, B., Siegel, L. S., & Wade-Woolley, L. (2001). Factors related to English reading performance in children with Chinese as a first language: More evidence of cross-language transfer of phonological processing. Journal of Educational Psychology, 93(3), 530–542. http://doi.org/10.1037//0022-0663.93.3.530 Gough, P. B., & Tunmer, W. E. (1986). Decoding, Reading, and Reading Disability. Remedial and Special Education, 7(1), 6–10. http://doi.org/10.1177/074193258600700104 Groff, P. (2001). Teaching phonics: Letter-to-phoneme, phoneme-to-letter, or both? Reading & Writing Quarterly, 17(4), 291–306. http://doi.org/10.1080/105735601317095043 Gunderson, L. (2014). ESL (ELL) Literacy Instruction A Guidebook to Theory and Practice (Thrid). New York: Routledge. Gunderson, L., & Siegel, L. S. (2001). The evils of the use define of IQ tests to define learning disabilities in first- and second- language learners. The Reading Teacher, 55(1), 48–55. Hamada, M., & Koda, K. (2008). Influence of first language orthographic experience on second language decoding and word learning. Language Learning, 58(1), 1–31. http://doi.org/10.1111/j.1467-9922.2007.00433.x Han, F. (2015). Word recognition research in foreign language reading : A systematic review. University of Sydney Papers in TESOL, 10, 57–91. Harm, M. W., & Seidenberg, M. S. (2004). Computing the meanings of words in reading: cooperative division of labor between visual and phonological processes. Psychological Review, 111(3), 662–720. http://doi.org/10.1037/0033-295X.111.3.662 Hatcher, P. J., Goetz, K., Snowling, M., Hulme, C., Gibbs, S., & Smith, G. (2006). Evidence for the effectiveness of the Early Literacy Support programme. The British Journal of Educational Psychology, 76, 351–367. http://doi.org/10.1348/000709905X39170 Hatcher, P. J., Hulme, C., Miles, J. N. V, Carroll, J. M., Hatcher, J., Gibbs, S., … Snowling, M. J. (2006). Efficacy of small group reading intervention for beginning readers with reading-delay: A randomised controlled trial. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 47(8), 820–7. http://doi.org/10.1111/j.1469-7610.2005.01559.x Hochberg, Y., & Benjamini, Y. (1990). More powerful procedures for multiple significance testing. Statistics in Medicine, 7, 811–818. Hoien, T., Lundberg, I., Stanovich, K., & Bjaalid, I. B. (1995). Components of phonological awareness. Reading and Writing: An Interdisciplinary Journal, 7, 171–188. Holm, A., & Dodd, B. (1996). The effect of first written language on the acquisition of English literacy. Cognition, 59(2), 119–147.  151  Hong, S. C., & Chen, S. H. (2011). Roles of position, stress, and proficiency in L2 children’s spelling: A developmental perspective. Reading and Writing, 24, 355–385. Hoover, W. A., & Gough, P. B. (1990). The simple view of reading. Reading and Writing, 2(2), 127–160. Howie, J. M. (1976). Acoustical studies of Mandarin vowels and tones. New York: Cambridge University Press. Hu, C. (2013). Predictors of reading in children with Chinese as a first language: A developmental and cross-linguistic perspective. Reading and Writing, 26(2), 163–187. http://doi.org/10.1007/s11145-012-9360-0 Hu, C., & Catts, H. W. (1998). The role of phonological processing in early reading ability: What we can learn from Chinese. Scientific Studies of Reading, 2(1), 55–79. http://doi.org/10.1207/s1532799xssr0201 Hu, C., & Schuele, C. M. (2015). When language experience fails to explain word reading development: Early cognitive and linguistic profiles of young foreign language learners. The Modern Language Journal, 99(4), 754–770. http://doi.org/10.1111/modl.12274 Hu, G. (2005). English language teaching in China: Policies, progress, and problems. Lanugage Policy, 24(4), 5–24. http://doi.org/10.1080/01434630308666503 Huang, C., Lin, S., & Su, S. (2004). The effects of phonological awareness training on technological university students’ phonics and vocabulary knowledge. Journal of National Taipei , 17(1), 59–90. Hulme, C., Bowyer-Crane, C., Carroll, J. M., Duff, F. J., & Snowling, M. J. (2012). The causal role of phoneme awareness and letter-sound knowledge in learning to read: Combining intervention studies with mediation analyses. Psychological Science, 23(6), 572–7. http://doi.org/10.1177/0956797611435921 Hulme, C., Caravolas, M., Málková, G., & Brigstocke, S. (2005). Phoneme isolation ability is not simply a consequence of letter-sound knowledge. Cognition, 97(1). http://doi.org/10.1016/j.cognition.2005.01.002 Hulme, C., Hatcher, P. J., Nation, K., Brown, A., Adams, J., & Stuart, G. (2002). Phoneme awareness is a better predictor of early reading skill than onset-rime awareness. Journal of Experimental Child Psychology, 82(1), 2–28. http://doi.org/10.1006/jecp.2002.2670 Jamshidian, M., Jalal, S., & Jansen, C. (2014). MissMech: An R package for testing homoscedasticity, multivariate normality, and missing completely at random (MCAR). Journal of Statistical Software, 56(6). Jiang, N. (2000). Lexical representation and development in a second language. Applied Linguistics, 47–77. Jiang, N. (2004). Semantic transfer and its implications for vocabulary teaching in a second language. The Modern Language Journal, 88, 416–432. Johnson, R. K., & Swain, M. (1997). Immersion in a foreign language. In Immersion education: International perspectives (pp. 6–8). Cambridge ; New York: Cambridge University Press. Johnston, R. S., Anderson, M., & Holligan, C. (1996). Knowledge of the alphabet and explicit awareness of phonemes in pre-readers: The nature of the relationship. Reading & Writing,  152  8(3), 217–234. http://doi.org/10.1007/BF00420276 Joseph, L. M. (1999). Developing first graders’ phonemic awareness, word identification and spelling: A comparison of two contemporary phonic instructional approaches. Reading Research and Instruction, 39(2), 160–169. http://doi.org/10.1080/19388070009558318 Kahn-Horwitz, J., Shimron, J., & Sparks, R. L. (2005). Predicting foreign language reading achievement in elementary school students. Reading and Writing, 18(6), 527–558. http://doi.org/10.1007/s11145-005-3179-x Kahn-Horwitz, J., Shimron, J., & Sparks, R. L. (2006). Weak and strong novice readers of English as a foreign language : Effects of first language and socioeconomic status. Annals of Dyslexia, 56(1), 161–185. Keung, Y. C., & Ho, C. S. H. (2009). Transfer of reading-related cognitive skills in learning to read Chinese (L1) and English (L2) among Chinese elementary school children. Contemporary Educational Psychology, 34(2), 103–112. http://doi.org/10.1016/j.cedpsych.2008.11.001 Kim, Y. S., Petscher, Y., Foorman, B. R., & Zhou, C. (2010). The contributions of phonological awareness and letter-name knowledge to letter-sound acquisition—a cross-classified multilevel model approach. Journal of Educational Psychology, 102(2), 313–326. http://doi.org/10.1037/a0018449 Knell, E. S. (2010). A longitudinal study of early English immersion and literacy in Xi’an, China. Retrieved from ProQuest Digital Dissertations. University of Utah, Salt Lake. Knell, E. S., & Chi, Y. (2012). The roles of motivation , affective attitudes , and willingness to communicate among Chinese students in early English immersion programs. International Education, 41(2). Knell, E. S., Haiyan, Q., Miao, P., Yanping, C., Siegel, L. S., Lin, Z., & Wei, Z. (2007). Early English immersion and literacy in Xi’an, China. The Modern Language Journal, 91(3), 395–417. Krashen, S. (2002). Defending whole language: The limits of phonics instruction and the efficacy of whole language instruction. Reading Improvement, 39(1), 32-42. Lai, Y., Tsai, H., & Yu, P. (2009). A multimedia English learning system using HMMs to improve phonemic awareness for English learning. International Forum of Educational Technology & Society, 12(3), 266–281. Lamb, M. (2011). A Matthew Effect in English language education in a developing country context. In Dreams and Realities:Developing Countries and the English Language (Vol. 20, pp. 186–206). http://doi.org/10.1016/j.aap.2006.08.007 Lau, C., & Rao, N. (2013). English vocabulary instruction in six early childhood classrooms in Hong Kong. Early Child Development and Care, 183(10), 1363–1380. http://doi.org/10.1080/03004430.2013.788815 Lee, J. A. C., & Al Otaiba, S. (2017). End-of-kindergarten spelling outcomes: How can spelling error analysis data inform beginning reading instruction? Reading & Writing Quarterly, 33(3), 226–238. http://doi.org/10.1080/10573569.2016.1165639 Lenhard, W. & Lenhard, A. (2016). Calculation of Effect Sizes. available:  153  https://www.psychometrica.de/effect_size.html. Dettelbach (Germany): Psychometrica. DOI: 10.13140/RG.2.1.3478.4245 Li, H., Rao, N., & Tse, S. K. (2012). Adapting western pedagogies for Chinese literacy instruction: Case studies of Hong Kong, Shenzhen, and Singapore preschools. Early Education & Development, 23(4), 603–621. http://doi.org/10.1080/10409289.2010.536441 Li, M., Kirby, J. R., Cheng, L., Wade-woolley, L., & Qiang, H. (2012). Cognitive predictors of English reading achievement in Chinese English-immersion students. Reading Psychology, 33(5), 423–447. http://doi.org/10.1080/02702711.2010.528663 Lin, F., & Cheng, C.-K. (2008). The combined effects of the keyword method and phonics instruction on 5 th Graders ’ English vocabulary learning in Taiwan. English Teaching & Learning, 3, 71–114. Liu, X., & Tao, S. (2007). The relation between vocabulary and reading in English learner of Mandarin-speaking Children. Acta Psychologica Sinica, 39, 118–128. McBride-Chang, C., Bialystok, E., Chong, K. K. Y., & Li, Y. (2004). Levels of phonological awareness in three cultures. Journal of Experimental Child Psychology, 89(2), 93–111. http://doi.org/10.1016/j.jecp.2004.05.001 McBride-Chang, C. (2004). Children’s literacy development. New York, NY: Oxford University Press. McDowell, H. J., & Lorch, M. P. (2008). Phonemic awareness in Chinese L1 readers of English: Not simply an effect of orthography. TESOL Quarterly, 42(3), 495-513. Melby-Lervåg, M., & Lervåg, A. (2011). Cross-linguistic transfer of oral language, decoding, phonological awareness and reading comprehension: A meta-analysis of the correlational evidence. Journal of Research in Reading, 34(1), 114–135. http://doi.org/10.1111/j.1467-9817.2010.01477.x Melby-Lervåg, M., Lyster, S.-A., & Hulme, C. (2012). Phonological skills and their role in learning to read: A meta-analytic review. Psychological Bulletin, 138(2), 322–352. http://doi.org/10.1037/a0026744 Metsala, J. (1999). Young children’s phonological awareness and nonword repetition as a function of vocabulary development. Journal of Educational Psychology, 91(1), 3–19. http://doi.org/10.1037/0022-0663.91.1.3 Metsala, J. & Walley, A. (1998). Spoken vocabulary growth and the segmental restructuring of lexical representations: Precursors to phonemic awareness and early reading ability. In J. Metsala & L. Ehri (Eds.), Word recognition in beginning literacy (pp. 89–120). Mahwah, NJ: Lawrence Erlbaum Associates. Ministry of Education P.R. China. (2011). Curriculum Standards for Primary School English Education. Muljani, D., Koda, K., & Moates, D. R. (1998). The development of word recognition in a second language. Applied Psycholinguistics, 19(1), 99-113. Muter, V., Hulme, C., Snowling, M., & Taylor, S. (1998). Segmentation, not rhyming, predicts early progress in learning to read. Journal of Experimental Child Psychology, 71(1), 3–27. http://doi.org/10.1006/jecp.1998.2453  154  Nation, K., & Snowling, M. J. (1998). Semantic processing and the development of word-recognition Skills : Evidence from children with reading comprehension difficulties, Journal of memory and language, 101(39), 85–101. Nishanimut, S. P., Johnston, R. S., Joshi, R. M., Thomas, P. J., & Padakannaya, P. (2013). Effect of synthetic phonics instruction on literacy skills in an ESL setting. Learning and Individual Differences, 27, 47–53. http://doi.org/10.1016/j.lindif.2013.06.007 Nunan, D. (2003). The impact of english as a global language on educational policies and practices in the Asia-Pacific region. TESOL Quarterly, 37(4), 589–613. http://doi.org/10.1080/13488678.1998.10800995 Ouellette, G. P. (2006). What’s meaning got to do with it: The role of vocabulary in word reading and reading comprehension. Journal of Educational Psychology, 98(3), 554–566. http://doi.org/10.1037/0022-0663.98.3.554 Perfetti, C. A., & Hart, L. (1991). The lexical bases of comprehension skills. In D. Gorphien (Ed.), On the consequences of meaning selection (pp. 67–86). Washington DC: American Psychological Association. Perfetti, C. A., & Hart, L. (2002). Lexical quality hypothesis. In L. Verhoeven, C. Elbro, & P. Reitsma (Eds.), Precursors of Functional literacy (pp. 190–213). Amsterdam/Philadelphia: John Benjamins Publishing Company. Qiang, H., & Siegel, L. S. (2012). Introduction of English immersion in China: A transplant with modifications. International Education, 41(2), 6-18. Richards, J. C., & Rodgers, T. S. (2014). Approaches and methods in language teaching. Cambridge university press. Rixon, S. (2011). Beyond ABC : Investigating current rationales and systems for the teaching of early reading to young learners of English. Retrieved from ProQuest Digital Dissertations. University of Warwick, Coventry. Saiz, M. (2007). Does l2 word decoding imply l2 meaning activation? Relationships among decoding, meaning identification, and l2 oral language proficiency in reading spanish as a second. Retrieved from ProQuest Digital Dissertations. University of Pittsburgh, Pittsburgh. Seidenberg, M. S., & McClelland, J. L. (1989). A distributed, developmental model of word recognition and naming. Psychological Review, 96(4), 523–68. Shafer, R. (1955). Classification of the Sino-Tibetan languages. WORD, 11(1), 94–111. Share, D. L. (1995). Phonological recoding and self-teaching: Sine qua non of reading acquisition. Cognition, 55(2), 151–218. Shiotsu, T. (2009). Reading ability and components of word recognition speed: The case of L1-Japanese EFL learners. In Second language reading research and instruction: Crossing the boundaries (pp. 15–39). Shu, H., Chen, X., Anderson, R. C., Wu, N., & Xuan, Y. (2003). Properties of school Chinese: Implications for learning to read. Child Development, 74(1), 27–47. http://doi.org/10.1111/1467-8624.00519 Shum, K. K., Ho, C. S.-H., Siegel, L. S., & Au, T. K. (2016). First-language longitudinal predictors of second-language literacy in young l2 learners. Reading Research Quarterly,  155  51(3), 323–344. http://doi.org/10.1002/rrq.139 Siegel, L. S. (2011). Early English immersion in Xi’an, China: An experiment in English language teaching. Frontiers of Education in China, 6(1), 1–7. http://doi.org/10.1007/s11516-011-0119-1 Siegel, L. S., & Ryan, E. B. (1989). The development of working memory in normally achieving and subtypes of learning disabled children. Child Development, 60, 973–980. Smith, F. (1977). Making sense of reading-and of reading instruction. Harvard Educational Review, 47(3), 386–395. Smith, F. (1999). Why systematic phonics and phonemic awareness instruction constitute an educational hazard. Language Arts, 77(2), 150-155. So, D., & Siegel, L. S. (1997). Learning to read Chinese: Semantic, syntactic, phonological and working memory skills in normally achieving and poor Chinese readers. Reading and Writing, 9(1), 1–21. http://doi.org/10.1023/A:1007963513853 Song, X., & Cheng, L. (2011). Investigating primary English immersion teachers in China: Background, instructional contexts, professional development, and perceptions. Asia-Pacific Journal of Teacher Education, 39(2), 97–112. http://doi.org/10.1080/1359866X.2011.560647 Sparks, R. L. (1995). Examining the linguistic coding differences hypothesis to explain individual differences in foreign language Learning. Annals of Dyslexia, 45, 187–214. Sparks, R. L. (2006). Is there a “Disability” for learning a foreign language? Journal of Learning Disabilities, 39(6), 544–557. http://doi.org/10.1177/00222194060390060601 Stanovich, K. E. (1986). Matthew Effects in reading : Some consequences of individual differences in the acquisition of literacy. Reading Research Quartely, 21(4), 360–407. Stuart, M. (1999). Getting ready for reading: Early phoneme awareness and phonics teaching improves reading and spelling in inner-city second language learners. British Journal of Educational Psychology, 69(4), 587–605. http://doi.org/10.1348/000709999157914 Tangel, D., & Blachman, B. (1992). Effect of phoneme awareness instruction on kindergarten children’s invented spelling. Journal of Literacy Research, 26(2), 233– 261. http://doi.org/10.1080/10862969509547876 Tangel, D., & Blachman, B. (1995). Effect of phoneme awareness instruction on the invented spelling of first-grade children: A one-year follow-up. Journal of Literacy Research, 27(2), 233– 261. http://doi.org/10.1080/10862969509547876 Team, R. C. (2013). R: A language and environment for statistical computing. Vienna, Austria.: R Foundation for Statistical Computing. Thompson, G., Cottrell, D., & Fletcher-Flinn, C. (1996). Sublexical orthographic-phonological relations early in the acquisition of reading: The knowledge sources account. Journal of Experimental Child Psychology, 62(2), 190–222. http://doi.org/10.1006/jecp.1996.0028 Thorius, K. K., & Sullivan, A. L. (2013). Interrogating instruction and intervention in RTI research with students identified as English language learners. Reading & Writing Quarterly, 29(1), 64–88. http://doi.org/10.1080/10573569.2013.741953 Treiman, R. (1991). Children ’ s spelling errors on syllable-initial consonant clusters. Journal of  156  Educational Psychology, 83(3), 346–360. Tunmer, W. E., Nesdale, A. R., & Wright, A. D. (1987). Syntactic awareness and reading acquisition. British Journal of Developmental Psychology, 5(1), 25–34. http://doi.org/10.1111/j.2044-835X.1987.tb01038.x Verbeke, G., & Molenberghs, G. (2009). Linear mixed models for longitudinal data. Springer Science & Business Media. Wagner, R. K., & Torgesen, J. K. (1987). The nature of phonological processing and its causal role in the acquisition of reading skills. Psychological Bulletin, 101(2), 192–212. Walley, A. C., Metsala, J. L., Victoria, M., & Garlock, V. M. (2003). Spoken vocabulary growth: Its role in the development of phoneme awareness and early reading ability. Reading and Writing, 16(1), 5–20. http://doi.org/10.1023/A:1021789804977 Wang, M., & Geva, E. (2003). Spelling acquisition of novel English phonemes in Chinese children. Reading and Writing, 16(4), 325–348. http://doi.org/10.1023/A:1023661927929 Wang, M., Perfetti, C. A, & Liu, Y. (2005). Chinese-English biliteracy acquisition: Cross-language and writing system transfer. Cognition, 97(1), 67–88. http://doi.org/10.1016/j.cognition.2004.10.001 Wolf, M., & Katzir-Cohen, T. (2001). Reading fluency and its intervention. Scientific studies of reading, 5(3), 211-239. Wu, X., Li, W., & Anderson, R. C. (1999). Reading instruction in China. Journal of Curriculum Studies, 31(5), 571–586. http://doi.org/10.1080/002202799183016 Ye, Z., & Zhou, X. (2009). Executive control in language processing. Neuroscience and Biobehavioral Reviews, 33(8), 1168–77. http://doi.org/10.1016/j.neubiorev.2009.03.003 Yeong, S. H. M., & Rickard Liow, S. J. (2011). Cognitive–linguistic foundations of early spelling development in bilinguals. Journal of Educational Psychology, 103(2), 470–488. http://doi.org/10.1037/a0022437 Yeung, S. S., & Chan, C. K. K. (2013). Phonological awareness and oral language proficiency in learning to read English among Chinese kindergarten children in Hong Kong. British Journal of Educational Psychology, 83(4), 550–568. http://doi.org/10.1111/j.2044-8279.2012.02082.x Yeung, S. S., Siegel, L. S., & Chan, C. K. K. (2013). Effects of a phonological awareness program on English reading and spelling among Hong Kong Chinese ESL children. Reading and Writing, 26(5), 681–704. http://doi.org/10.1007/s11145-012-9383-6 Yin, L., Anderson, R. C., & Zhu, J. (2007). Stages in Chinese children’s reading of English words. Journal of Educational Psychology, 99(4), 852–866. http://doi.org/10.1037/0022-0663.99.4.852 Zhang, D. (2012). Chinese primary school English curriculum reform. In Perspectives on teaching and learning English literacy in China (pp. 67–83). Springer Netherlands. Zhang, Z. (1993). 张志公语言和语文教育思想硏讨会论文选集. 语文出版社. Zhou, Y., & Mcbride-chang, C. (2009). Teaching English as a foreign language to young Chinese children. Journal of Communications Research, 3(1), 103–115.  157  Ziegler, J. C., & Goswami, U. (2005). Reading acquisition, developmental dyslexia, and skilled reading across languages: A psycholinguistic grain size theory. Psychological Bulletin, 131(1), 3–29. http://doi.org/10.1037/0033-2909.131.1.3    158  APPENDICES Appendix A Sample lesson in phonics  I. Whole word recognition     II. Chant for word decoding e.g. bus “B” “B” “B” ,/b/, /b/,/ b/ ;”U” /^/,/^/,/^/,; /b/ /^/, /b^/, “S” /s/,/s/,/s/, /b^S/. III. Decodable text  IV. Letter sound correspondences, reading the story and again and pay much attention to the sound of each letter that you fill in the following blanks  159   V. Discrimination of upper and lower cases. Match the words with pictures and read them aloud     160  Appendix B English learning experience survey  Dear parents/guardians: We are researchers from University of British Columbia, Canada. This term, we are carrying out a research project in your child’s school investigating teaching and learning English among Chinese teachers and students. In order to gather information of your child learning English after school, we invite you to the fill in the following questionnaire. The information you provide will be confidential and not be used for other purposes. Thank you for your cooperation!                                      Xi’an English literacy research team                                                           2015-8-14 A. Background information  Child’s name:             Gender:         Grade:______             Date of birth:      YYYY     MM     DD Your relation to the child (please put a‘√’on the correct number) :①Child’s father ② child’s mother③ child’s guardian other than parents  B. Early experience of your child learning English 1. At what age did your child start to learn English? ____ Year___ Month 2. In what ways did your child learn English at the beginning?(please select all items that apply and put a ‘√’ on the correct number): ①taught by parents or other guardians (grandparents)  ②went to English classes at a private institute ③ taught by a private tutor ④ learned English lessons at school ⑤other, please specify _________________________________  C. English learning at home 3. Currently, do you or other guardians teach your child English at home?  ①Yes__ ②No___ If yes, what do you teach him/her? (please select all items that apply and put a ‘√’ on the correct number)  161  ①to read English alphabet;②to say English words; ③to say simple English sentences; ④ to write English alphabet; ⑤to spell English words; ⑥to write simple English sentences; ⑦ to sing English songs or rhyme nurseries 4. Currently, how often do you teach your child English?  ①never; ②once a month or less; ③2 to 3 times a month; ④ once a week; ⑤2 to 3 times a week; ⑥once a day or more 5. In general, how much time do you spent on teaching your child English one time? ①never;②less than 10 minutes;③10-20 minutes;④20-30 minutes; ⑤more than 30 minutes 6. Please evaluate frequency of the following activities you child engage at home. [put a ‘√’ on the correct number ] The numbers represent: ①      ② ③ ④ ⑤ ⑥ never once a month or less 2 to 3 times a month once a week 2 to 3 times a week once a day or more  Listen to or read English stories            ① ② ③ ④ ⑤ ⑥ Watch English videos          ① ② ③ ④ ⑤ ⑥ Watch English television program ① ② ③ ④ ⑤ ⑥ Learn to say English words or sentences         ① ② ③ ④ ⑤ ⑥ Listen to or sing English songs             ① ② ③ ④ ⑤ ⑥ Play games in English on digital devices              ① ② ③ ④ ⑤ ⑥  D. English learning from extra curriculum activities 7. Has your child ever participated or is participating extracurricular activities to learn English (eg. English club, private lessons)  ①Yes__ ②No__  162  If yes, please specify names of the activities and how much time spent on the activities?     Time                         Name         Frequency            Duration ____Year___Month to____Year___Month; ________;    ______times/week;   _____hours/time ____Year___Month to ____Year___Month; ________;    ______times/week;   _____hours/time ____Year___Month to ____Year___Month; ________;    ______times/week;   _____hours/time 8. Have you ever hired an English tutor for your child?    ①Yes____  ②No____ If yes,please specify when, times per week and hours per time.  Time                                     Frequency             Duration   ____Year_____Month to ____Year___Month;     ______times/week;   _____hours/time; ____Year_____Month to ____Year___Month;     ______times/week;   _____hours/time; ____Year_____Month to ____Year___Month;     ______times/week;   _____hours/time;  E. English language and print exposure at home 9. How many books do you have at your home about children learning English? ________          Please provide at least one example:                                                         10. How many kinds of magazine and papers do you order for your child to learn English?   ________     Please provide at least one example:                                                         11. How many kinds of audio and video material do you have at home for your child to learn English?(e.g. apps on your phone, DVDs)?  ________     Please provide at least one example:                                                         12. How many kinds of games and toys do you have at home for your child to learn English (e.g. letter cubes, flash cards)?_______     Please provide at least one example:                                                   163  13. How many kinds of toys do you have at home for your child to learn English (e.g. letter cubes, flash cards)?_______     Please provide at least one example:                                                  14. How many apps do you have on your digital device for your child to learn English (e.g. Phonics Monster)?_______ Please provide at least one example:                                                  F. Parents’ education and English proficiency level  15. The highest level of education the child’s father received is:        ①Primary school;②middle school;③secondary school;④vocational secondary school;junior college;⑤undergraduate;⑥graduate and above       16. The highest level of education the child’s mother received is:        ①Primary school;②middle school;③secondary school;④vocational secondary school;junior college;⑤undergraduate;⑥graduate and above           164  Appendix C Chinese character identification        东 牛 儿 朋 左 星 早 爷 十 热 共 事 开 的 爸 我 山 得 过 狗 理 兴 班 华 掉 呢     遇 渴   能 拔 舌 习 喝 急 摇 答 瓶 位 聪 凉 承 嗓 蜓 备 塞 孤 粱 蛾 糖 芹 莺 揪 菠 跨 察 所 含 灿 交 混 库 刹 牢 绝 泛 叛 扒 肤 凿 洒 罗 谅 梯 固 杏 繁 统 甩 菌 舱 妙 怜 硫 哎 磷 枫 奥 灶 柄 哼 坟 丘 帜 瑰 玲 骚 妄 嘀 嘹 疤 苔 愚 埃 橄 诸 蓉 砰 弥 咙 缰 贷 棺 豌 啼 啤 侄 崩 椰 狈 蔼 贞 肛 卢 搀 韭 碾 肪 阀 澳 疙 俺 浊 茵 肆 畔 拂 蚀 晤 舵       165  Appendix D Pinyin spelling items             1.bǎ 2.hú 3.cì 4. dé 5. shā 6. nián 7.huàn 8.jiào 9.wǒ 10.yù 11.lè 12.zhù 13.guǎn 14.zhuāng 15.chéng 16.shuài 17.chí  18.fǒu 19.bǎi 20.sōu 21.kuò 22.zéi 23.hěn 24.ān  166  Appendix E Oral cloze items        167  Appendix F Picture naming items   168  Appendix G Word recognation tasks  I. Reading words aloud  II. Word meaning access 1 cat a. 猫 b. 狗 c. 可爱 d. 不认识 2 my a. 你的 b. 我的 c. 他的 d. 不认识 3 red a. 颜料 b. 红色 c. 蓝色 d. 不认识 4 yes a. 否 b. 也许 c. 是 d. 不认识 5 to a. 到,表示目的 b. 数字二 c. 从…来 d. 不认识 6 no a. 可能 b. 不 c. 是 d. 不认识 7 me a. 你的 b. 它 c. 我 d. 不认识 8 car a. 自行车 b. 艺术 c. 汽车 d. 不认识 9 see a. 看到 b. 听到 c. 蜜蜂 d. 不认识 10 it a. 它 b. 他们的 c. 是 d. 不认识 11 up a. 进去 b. 上 c. 下 d. 不认识 12 we a. 我们 b. 他们 c. 然后 d. 不认识 13 he a. 这 b. 她 c. 他 d. 不认识 14 the a. 表示特定的一个 b. 一个 c. 多个 d. 不认识 15 is a. 是 b. 不是 c. 可能是 d. 不认识 16 and a. 但是 b. 然后 c. 和 d. 不认识 17 go a. 来 b. 去 c. 哎呦 d. 不认识 18 will a. 表示将来 b. 表示过去 c. 表示现在 d. 不认识 19 not a. 也许是 b. 不是 c. 但是 d. 不认识 20 but a. 然后 b. 这 c. 但是 d. 不认识 21 from a. 从…来 b. 上面 c. 下面 d. 不认识 22 had a. 是 b. 过去有 c. 过去在 d. 不认识 23 on a. 在…下面 b. 在…前面 c. 在…上面 d. 不认识 24 of a. 表示谁的 b. 表示在哪 c. 表示时间 d. 不认识  169  25 him a. 他 b. 她 c. 我 d. 不认识 26 as a. 谁的 b. 作为 c. 过去 d. 不认识 27 book a. 桌子 b. 笔 c. 书 d. 不认识 28 was a. 表示将来 b. 表示过去 c. 表示现在 d. 不认识 29help a. 帮助 b. 爱 c. 问题 d. 不认识 30 then a. 未来 b. 比较 c. 当时,那是 d. 不认识 31 time a. 空间 b. 未来 c. 时间 d. 不认识 32 let a. 制作 b. 让 c. 引导 d. 不认识 33 men a. 男人 b. 孩子 c. 女人 d. 不认识 34 baby a. 巨人 b. 婴儿、宝贝 c. 洗澡 d. 不认识 35 new a. 未来的 b. 旧的 c. 新的  d. 不认识 36 stop a. 停止 b. 继续 c. 通过 d. 不认识 37 work a. 工作 b. 娱乐 c. 休息 d. 不认识 38 part a. 一点点 b. 部分 c. 整体 d. 不认识 39 fast a. 慢 b. 快 c. 高 d. 不认识 40 fine a. 生病 b. 假的 c. 很好的 d. 不认识 41 milk a. 牛奶 b. 做 c. 奶牛 d. 不认识 42 keep a. 借阅 b. 保持,继续 c. 钥匙 d. 不认识 43 said a. 过去唱 b. 过去走路 c. 过去说,据说 d. 不认识 44 with a. 谢谢 b. 过去 c. 用,和 d. 不认识 45 light a. 光 b. 火 c. 力量 d. 不认识 46 their a. 我们的 b. 他们的 c. 你们的 d. 不认识 47 which a. 在哪 b. 哪一个 c. 谁 d. 不认识 48 would a. 表示会 b. 表示不会 c. 表示拒绝 d. 不认识 49 use a. 借 b. 使用 c. 出租 d. 不认识 50 together a. 见面 b. 分别 c. 一起 d. 不认识       170  Appendix H Pseudoword reading items      Simple structured words ip      ga     ko    ta    om      nat   ib    hap    mell    fim    ven   jop    pim     wum    lat   baf     din    nup    fet   i. Words with complex patterns floxy      leck   pawk  distrum   chur     vorse      gradly loast    blighten    wreet    yerdle     koodoo     baunted  171  Appendix I Rhyme detection items  Demonstrating items  Test items 1.house                                                                horse      mouse      door 2.boat                                                                   goat        bike           foot 3.egg                                                                     leg           bag            spoon 4.bee                                                                     bear        cow            tree 5.chair                                                                   hair         table           car 6.head                                                                   hand       eye             bed 7.bell                                                                     box          dress          cell 8.sock                                                                   clothes      shoe         clock 9.train                                                                   truck         rain           spoon              10.car                                                                    cake           star            bike  172  Appendix J Phonemic awareness tasks  I. Initial phoneme deletion items Demonstrating items: 1) bus, 2) sad Test Items: 1) bear;  2) sit; 3) jam; 4)seat;  5) hat;  6) tin; 7) Cake II. Final phoneme deletion items 1) seat;  2) sad;  3)  hat;  4 ) jam;   5 ) tin III. Phoneme counting items Demonstrating item: (3) a.  (   Tom  )                                 b. (  pig    )         c. (  Zoo )                    d.                                   (  blue ) e. (Daddy )                      f.    (   egg ) g. ( glad )蓝色  173  Appendix K Fixed effects of English real word reading    Estimate Std. Error df t value p p (adjusted) (Intercept) 7.37 1.25 11.34 5.91 <.001 < .001 Time3 3.00 0.76 192.81 3.98 < .001 < .001 Time1 -5.75 0.72 189.85 -7.97 < .001 < .001 pri1n01 5.02 1.50 12.06 3.34 .01 .008 exCurDl 1.67 1.06 113.47 1.58 .12 .117 Grade3 6.07 1.40 8.09 4.33 .00 .005 Time3:pri1n01 0.52 1.07 196.22 0.49 .63 .627 Time1:pri1n01 2.34 0.96 191.01 2.45 .02 .04   174  Appendix L Fixed effects in model of pseudoword reading   Estimate Std. Error df t value p p (adjusted) (Intercept) 2.10 0.75 18.43 2.80 .01 .003 Time3 4.55 0.74 191.63 6.15 < .001 <.001 Time1 -1.95 0.68 183.74 -2.87 < .001 .009 pri1n01 1.61 0.92 20.91 1.75 .09 .07 exCurDl 0.58 0.64 116.62 0.90 .37 .36 Grade3 3.47 0.75 8.41 4.62 < .001 .002 Time3:pri1n01 2.11 1.04 201.24 2.03 .04 .03 Time1:pri1n01 -0.12 0.90 185.45 -0.13 .89 .89   175  Appendix M Fixed effects in model of simple pseudoword reading  Estimate Std. Error df t value p p(adjusted) (Intercept) 6.10 0.95 13.47 6.39 < .001 < .001 Time3 2.35 0.85 72.11 2.77 .01 .002 pri1n01 3.03 1.14 13.88 2.65 .02 .03 exCurDl 0.60 0.96 104.57 0.62 .53 .53 Grade3 3.06 1.03 8.40 2.96 .02 .03 Time3:pri1n01 -0.47 1.21 82.03 -0.39 .70 .72     176  Appendix N Fixed effects in model of reading pseudowords with complex patterns  Estimate Std. Error df t value p p (adjusted) (Intercept) 0.30 0.46 12.17 0.67 .52 .52 Time3 0.40 0.40 81.97 1.01 .31 .17 pri1n01 0.23 0.55 12.70 0.41 .69 .69 exCurDl 0.28 0.46 111.35 0.61 .55 .55 Grade3 1.47 0.50 7.81 2.94 .02 .04 Time3:pri1n01 1.46 0.57 90.93 2.54 .01 .02   177  Appendix O Fixed effects in model of initial phoneme deletion   Estimate Std. Error df t value p p (adjusted) (Intercept) 2.03 0.39 18.45 5.20 < .001 <.001 Time1 -0.79 0.44 96.56 -1.82 .07 . 04 pri1n01 1.04 0.49 22.83 2.11 .05 .03 exCurDl 0.47 0.36 101.43 1.30 .20 .20 Grade3 2.06 0.38 7.61 5.44 < .001 <.001 Time1:pri1n01 -0.35 0.59 100.83 -0.60 .55 .55   178  Appendix P Fixed effects in model of rhyme awareness  Estimate Std. Error df t value p p(adjusted) (Intercept) 5.40 0.37 255.88 14.75 <.001 <.001 Time3 0.22 0.43 188.24 0.52 .60 .40 Time1 -0.45 0.39 167.58 -1.16 .25 .35 pri1n01 1.04 0.44 260.71 2.38 .02 .02 exCurDl -0.52 0.33 98.68 -1.58 .12 .14 Grade3 0.98 0.29 95.61 3.37 <.001 <.001 Time3:pri1n01 0.02 0.57 180.96 0.03 .97 .97 Time1:pri1n01 -0.37 0.51 169.37 -0.72 .47 .47      179  Appendix Q Fixed effects in model of picture naming  Estimate Std. Error df t value p p (adjusted) (Intercept) 6.32 0.58 9.97 10.96 < .001 < .001 Time1 -2.45 0.26 101.16 -9.45 < .001 < .001 pri1n01 2.52 0.69 10.35 3.65 < .001 < .001 exCurDl 1.40 0.51 113.16 2.75 .01 .02 Grade3 2.04 0.66 7.81 3.08 .02 .03 Time1:pri1n01 0.27 0.35 102.00 0.79 .43 .43   180  Appendix R Fixed effects in model of oral cloze  Estimate Std. Error df t value p p(adjusted) (Intercept) 2.90 0.45 12.50 6.42 < .001 < .001 Time1 -0.45 0.32 92.44 -1.41 .16 .09 pri1n01 2.97 0.49 122.17 6.00 < .001 < .001 exCurDl 0.22 0.44 105.70 0.51 .61 .40 Grade3 1.32 0.52 2.93 2.52 .09 .05 Time1:pri1n01 -0.83 0.43 92.83 -1.93 .06 .08   181  Appendix S Fixed effects in model of Chinese character identification  Estimate Std. Error df t value p p (adjusted) (Intercept) 65.11 3.69 11.19 17.64 < .001 < .001 Time1 -14.38 1.87 92.09 -7.69 < .001 < .001 pri1n01 2.79 4.40 11.30 0.63 .54 0.54 exCurDl 8.34 4.02 105.57 2.07 .04 .04 Grade3 36.04 4.09 7.53 8.82 < .001 < .001 Time1:pri1n01 1.52 2.48 92.29 0.61 .54 .60   182  Appendix T Partial correlation of all the outcome variables at three time points controlling for age in study 1    1 2 3 4 5 6 7 8 9 10 11  Time 1                      1 Pypre -           2 Chrpre .16           3 Vocpre .17 .44**  .        4 Lnpre .22 .33** .54**         5 Widpre .27* .57** .73** .53**        6 Pswidpre .18 .51** .54** .34** .70**       7 Rhymepre .12 .34** .40** .33* .40** .41**      8 Ocpre .32* .39** .74** .49** .65** .52** .43**  .   9 intdepre .27* .16 .16 .21* .38** .31* .22* .30*    10 findepre .02 .11 .06 .12 .16 .11 .04 .14 .18   11 intidepre .12 .10 .03 .08 .19* .01 .08 .11 .11 .10   Time2                      13 vocpost .19* .41** .84** .60** .75** .47** .22* .67** .29* .08 .06 14 Lnpost .33* .33** .43** .49** .36** .24* .17 .41** .10 -.05 -.01 15 widpost .18 .46** .63** .48** .71** .53** .37** .51** .23* .08 .12 16 pswidpost .18 .39** .40** .27* .43** .40** .37** .39** .22* .13 .11 17 rhymepost .38* .33* .33* .29* .37** .38** .50** .38** .23* .08 .10 18 Ocpost .28* .43** .76** .49** .73** .47** .31** .73** .25* -.07 .12 19 intdepost .18* .39* .46** .34* .48** .42** .34** .37** .11 -.02 .05 20 findepost -.09 .06 .05 -.08 .02 -.08 .20 -.16 -.10 .10 .09 21 intidepost .13 .26* .31* .42* .33 .22 .18 .27* .14 -.02 .07 22 Sppost .21* .41** .56** .23* .49** .44** .26* .46** .17 .01 -.03  Time 3                      23 widDl .27* .50** .62** .48** .74** .64** .40** .51** .30* .10 .11 24 pswidDl .22 .50** .48** .15 .54** .59** .35** .31** .16 .15 .28* 25 spDl .30** .34** .43** .33* .49** .45** .29* .36** .29* .16 .13 Note. Pypre, Pinyin spelling pretest; Chrpre, Chinese character identification pretest; Vocpre, vocabulary (picture naming) pretest; Lnpre, letter naming pretest; Widpre, word identification pretest; Pswidpre, non word reading pretest; Rhymepre, rhyme awareness pretest; Ocpre, oral cloze pretest; intdepre, initial phoneme deletion pretest; findepre, final phoneme deletion pretest; intidepre, initial phoneme identification pretest; vocpost, vocabulary posttest; Lnpost, letter naming posttest; widpost, word identification posttest; pswidpost, non word reading posttest; rhymepost, rhyme awareness posttest; Ocpost, oral cloze posttest;  intdepost, initial phoneme deletion posttest; findepost, final deletion posttest; intidepost, initial phoneme identification posttest; Sppost, spelling posttest; widDl, word identification delayed test; pswidDl, non word reading delayed test; spDl, spelling delayed tests  183  Appendix S Partial correlation of all the outcome variable at three time points controlling for age continued    12 13 14 15 16 17 18 19 20 21  Time 2           12 Chrpost                    13 Vocpost .42**                  14 Lnpost .45** .48**                15 Widpost .48** .65** .28              16 Pswidpost .34** .39* .23 .71**            17 Rhymepost .29* .27 .23 .41* .35*          18 Ocpost .35* .74** .53** .57** .49* .30        19 Intdepost .18 .43* .20 .51** .36 .25 .46*      20 Findepost -.01 -.01 -.08 .04 .20 .20 .05 .09    21 Intidepost .29 .43 .30 .36* .27 .15 .44* .25 .07  22 Sppost .34** .64 .23 .49** .35* .44* .58** .52** .15 .34*  Time3           23 wroDl .38* .61** .17 .74** .50** .53** .61** .56** -.01 .33* 24 waDl .37* .39* -.01 .58** .57** .43* .47* .44* .16 .30* 25 SpDl .30* .45* .17 .45* .38* .49* .47* .57** .18 .35* Note. vocpost, vocabulary posttest; Lnpost, letter naming posttest; widpost, word identification posttest; pswidpost, non word reading posttest; rhymepost, rhyme awareness posttest; Ocpost, oral cloze posttest;  intdepost, initial phoneme deletion posttest; findepost, final deletion posttest; intidepost, initial phoneme identification posttest; Sppost, spelling posttest; widDl, word identification delayed test; pswidDl, non word reading delayed test; spDl, spelling delayed tests; *, p < .05 **, p <. 001.              184  Appendix U Chinese reading and spelling activities  Pinyin decoding, sounding out Pinyin scripts of words “Huo3 Che1”(Chinese word :火车, meaning train). Stroke sequencing, students were taught sequence of strokes to form a character. For example, the character “山” , meaning mountain, is written  in the following sequence,  Character copying, students were required to copy a novel character multiple times.  Word cloze, students were required to give a character to form a word with the given character.  Character structure awareness, students were taught to identify and discriminate single and compound characters, and to identify the pictographic and the phonetic element in a compound character.  185  Appendix V  Vocabulary list of the experimental program Compound words Schoolbag, basketball, blackboard, snowman, pencil-box Rhyming pairs and word families King, ring, sing Hat, rat, mat, cat Box, fox, socks Boy, toy Key, tree, tea, bee Two phoneme words Egg , ice, shoe, bee, arm, key, tree Three phoneme words Dog, cat, mat, cat, bed, mouth, sit, foot, nose, bus, cake, pig Four phoneme words Tiger, cookie, desk, baby, clock, bread Multi-syllabic words Elephant, hamburger, vegetable   186  Appendix W  Sample lesson plan of PA instruction Instructional goals. Identify words with 2 phonemes, segment words with 2 phonemes, blend phonemes into words Description. Phonemes are the smallest unit to represent sound in English language. For example, the word “cat” has three phonemes/k/ /ae/ /t/. Phonemic awareness refers to the knowledge that words consist of phonemes and the ability of manipulating phonemes in a word, such as segmenting, deleting and removing. Activity 1: Vocabulary (words of two phonemes) Teacher projects the pictures one by one on the screen, says the word out loud and says it again in a sentence provided below. Students repeat. To make it engaging to the students, teachers can demonstrate the word meaning with body movements. For example, teachers can pretend to eat an egg while saying the sentence.   Egg,  I like to eat eggs for breakfast. Shoe, I lost a shoe from running yesterday. Ice, Ice is made of water. Activity 2: Segmenting words into phonemes Instruction: “Let’s clap out our words” Teachers raise the right hand in front of the chest and pronounce /e/, and raise the left hand and pronounces /g/. Put two hands together and say “egg”. Students repeat after the teacher. Activity 3: Identify the words using phonemes The purpose of this activity is for students to identify the phonemes in words and review the words they have learnt. Split a picture of an egg in two parts. Show students the first half in the right hand and pronounce /e/. Ask student what word it is. Do the activity using at least four words.    187  Appendix X Spelling scoring criteria  CVC/VC words rat bed Car arm 0random string of letters     1.Single-letter response that represents some salient part of the world other than the initial phoneme. May be followed by a random string T/ Thudnn/a/e  d/e r r/m 2. The correct initial phoneme of the word. May be followed by a random string or an alphabet string Omission of a vowel r b C/k a 3. More than one phoneme(2), but not all. Must be represented with phonetically related or conventional letters. May include intrusions. When the intrusion is removed, the rest of the letters should be in proper sequence. Or represented all three phonemes, but the vowel is not correctly represented. Ra Ret/rua/ret/rut Be/ed/bad Ar /ca/ce/ka Rm/ um  4. Representation of all 3 phonemes with a mix of phonetically related and conventional letters. May include intrusions. Ruat/rate Bade/baid Care/kar Urm/erm/ 5. The correct spelling  rat bed        188  Appendix X spelling scoring criteria continued CCVC/CVCC Desk glad 0random string of letters   1.Single-letter response that represents some salient part of the world other than the initial phoneme. May be followed by a random string s/k l/d/a 2. The correct initial phoneme of the word. May be followed by a random string or an alphabetic string.  Or the vowel is also represented but the incorrect ones d da/du/ g/ge 3. More than one phoneme (2), but not all. Must be represented with phonetically related or conventional letters. May include intrusions. When the intrusion is removed, the rest of the letters should be in proper sequence. dek dsk gad/ gld/ga led/ad’/ 4. Representation of all phonemes with a mix of phonetically related and conventional letters. May include intrusions. Or may have the wrong order  daks/deks geld gald 5. All consonant phonemes, including initial blend, with conventional letters, the correct vowel, but the vowel is either not marked or marked incorrectly dask gled/glaid/glade 6. The correct spelling    189  Appendix Y outline of the Pinyin program and a sample lesson     Appendix Z Correlation coefficients of the outcomes controlling for age in study 2 1     1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 chrDl                                  2 vocDl 0.47                 3 lnDl 0.13 0.29                4 lsDl 0.19 0.37 0.51               5 widDl 0.56 0.62 0.36 0.46              6 pswidDl 0.38 0.46 0.19 0.33 0.57             7 rhymed 0.28 0.32 -0.04 0.1 0.25 0.16            8 ocDl 0.37 0.45 0.38 0.3 0.34 0.24 0.24           9 intdeDl 0.26 0.39 0.22 0.22 0.51 0.34 0.04 0.18          10 findeDl 0.14 0.1 0.36 0.19 0.15 0.16 -0.12 0.22 0.31         11 intideDl -0.06 0.02 0.07 0.18 -0.01 -0.05 -0.12 0 0.14 0.15        12 spDl 0.45 0.41 0.31 0.36 0.66 0.3 0.14 0.32 0.36 0.25 0.04       13 Pypre 0.22 0.45 0.12 0.08 0.39 0.29 0.19 0.19 0.23 0.01 0.01 0.17      14 Chrpre 0.79 0.32 0.07 0.03 0.37 0.3 0.19 0.23 0.21 0.26 -0.01 0.33 0.27     15 Vocpre -0.01 0.5 0.37 0.4 0.4 0.37 0.09 0.33 0.29 0.09 0.12 0.17 0.29 -0.02    16 Lnpre 0.12 0.35 0.52 0.52 0.4 0.24 0.13 0.39 0.09 0.27 0.12 0.35 0.13 0.18 0.58   17 Widpre 0.27 0.41 0.37 0.34 0.59 0.46 0.14 0.29 0.32 0.19 0.07 0.48 0.24 0.3 0.56 0.65  18 Pswidpre 0.13 0.32 0.34 0.35 0.29 0.38 0 0.3 0.04 0.23 0.06 0.22 0.2 0.15 0.47 0.62 0.64 Note. Pypre, Pinyin spelling pretest; Chrpre, Chinese character identification pretest; Vocpre, vocabulary (picture naming) pretest; Lnpre, letter naming pretest; Widpre, word identification pretest; Pswidpre, pseudoword reading pretest; Rhymepre, rhyme awareness pretest; Ocpre, oral cloze pretest; intdepre, initial phoneme deletion pretest; findepre, final phoneme deletion pretest; intidepre, initial phoneme identification pretest; vocpost, vocabulary posttest; Lnpost, letter naming posttest; widpost, word identification posttest; pswidpost, pseudoword reading posttest; rhymepost, rhyme awareness posttest; Ocpost, oral cloze posttest;  intdepost, initial phoneme deletion posttest; findepost, final deletion posttest; intidepost, initial phoneme identification posttest; Sppost, spelling posttest; chrpost, Chinese character identification posttest; pypost, Pinyin spelling, posttest;  widDl, word identification delayed test; lsDl, letter sound knowledge delayed test; pswidDl, non word reading delayed test; spDl, spelling delayed test.    191  Appendix Z Correlation coefficients of the outcomes controlling for age continued 2   1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Sylidpre 0.22 0.21 0.25 0.32 0.25 0.15 0.11 0.18 0.3 0.04 0.04 0.15 0.17 0.15 0.19 0.18 0.13 0.25 20 Phidpre 0.33 0.35 0.24 0.3 0.42 0.14 0.16 0.17 0.33 -0.08 0.18 0.31 0.2 0.17 0.26 0.18 0.25 0.14 21 rhymepre -0.04 0.02 -0.07 -0.31 -0.03 -0.06 0.05 0.02 0.02 -0.11 0.04 -0.05 0.16 0.14 0.15 0.02 0.05 0.06 22 Ocpre -0.04 0.21 0.53 0.24 0.26 0.17 -0.02 0.31 0.14 0.19 0.13 0.19 0.11 0.00 0.53 0.45 0.40 0.34 23 Intdepre 0.14 0.16 0.19 0.29 0.28 0.32 0.00 0.12 0.35 0.18 -0.04 0.37 0.25 0.18 0.23 0.31 0.45 0.38 24 Findepre 0.00 -0.02 0.35 -0.06 0.17 0.03 -0.06 0.09 0.19 0.25 0.01 0.26 0.25 0.13 0.11 0.20 0.23 0.19 25 intidepre 0.16 0.16 0.02 0.07 0.07 0.03 0.15 0.09 0.06 -0.02 0.01 0.20 -0.04 0.20 0.01 0.16 0.20 0.16 26 Chrpost 0.84 0.30 0.12 0.08 0.39 0.30 0.19 0.29 0.20 0.22 -0.14 0.35 0.27 0.90 -0.01 0.19 0.34 0.21 27 Vocpost 0.11 0.56 0.53 0.51 0.47 0.32 0.19 0.42 0.31 0.28 0.14 0.31 0.23 0.10 0.86 0.69 0.56 0.50 28 Lnpost 0.10 0.39 0.63 0.54 0.48 0.28 0.06 0.41 0.19 0.23 0.12 0.40 0.18 0.17 0.58 0.91 0.68 0.56 29 Widpost 0.44 0.36 0.40 0.46 0.64 0.40 0.12 0.43 0.33 0.27 0.06 0.62 0.25 0.39 0.37 0.60 0.70 0.55 30 pswidpost 0.33 0.32 0.28 0.37 0.53 0.55 0.00 0.41 0.38 0.28 -0.02 0.46 0.29 0.25 0.33 0.45 0.53 0.54 31 rhymepost 0.22 0.23 0.21 0.06 0.17 0.08 0.10 0.15 0.38 0.08 -0.08 0.22 0.16 0.04 0.28 0.09 0.07 0.02 32 Ocpost 0.09 0.50 0.37 0.36 0.36 0.19 0.13 0.34 0.27 0.16 0.14 0.29 0.07 0.03 0.63 0.53 0.44 0.38 33 intdepost 0.13 0.27 0.22 0.32 0.36 0.35 0.10 0.32 0.45 0.24 0.07 0.41 0.25 0.10 0.31 0.25 0.31 0.28 34 findepost 0.24 0.27 0.22 0.28 0.38 0.21 0.35 0.37 0.23 0.23 0.21 0.46 0.24 0.23 0.16 0.24 0.21 0.13 35 intidepost -0.19 -0.10 -0.04 -0.09 -0.16 0.05 -0.05 0.01 0.13 -0.01 0.16 0.03 -0.14 -0.13 -0.03 -0.08 0.11 -0.10 36 Pypost 0.08 0.30 0.01 0.23 0.33 0.15 0.12 0.19 0.20 -0.03 0.05 0.20 0.35 0.07 0.24 0.08 0.17 0.10 Note. Pypre, Pinyin spelling pretest; Chrpre, Chinese character identification pretest; Vocpre, vocabulary (picture naming) pretest; Lnpre, letter naming pretest; Widpre, word identification pretest; Pswidpre, non word reading pretest; Rhymepre, rhyme awareness pretest; Ocpre, oral cloze pretest; intdepre, initial phoneme deletion pretest; findepre, final phoneme deletion pretest; intidepre, initial phoneme identification pretest; vocpost, vocabulary posttest; Lnpost, letter naming posttest; widpost, word identification posttest; pswidpost, non word reading posttest; rhymepost, rhyme awareness posttest; Ocpost, oral cloze posttest;  intdepost, initial phoneme deletion posttest; findepost, final deletion posttest; intidepost, initial phoneme identification posttest; Sppost, spelling posttest; chrpost, Chinese character identification posttest; pypost, Pinyin spelling, posttest;  widDl, word identification delayed test; lsDl, letter sound knowledge delayed test; pswidDl, non word reading delayed test; spDl, spelling delayed test.    192   Appendix Z Correlation coefficients of the outcomes controlling for age continued 3   19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 21 rhymepre 0.08 -0.09                               22 ocpre -0.03 0.06 0.07               23 intdepre 0.16 0.29 -0.03 0.17              24 findepre 0.22 0.32 0.11 0.17 0.37             25 intidepre 0.04 0.10 0.08 0.06 0.18 0.05            26 chrpost 0.18 0.28 0.06 -0.03 0.22 0.12 0.20           27 vocpost 0.29 0.29 0.11 0.54 0.24 0.21 0.09 0.08          28 lnpost 0.22 0.26 0.05 0.48 0.35 0.20 0.14 0.18 0.67         29 widpost 0.16 0.36 -0.01 0.37 0.48 0.24 0.19 0.50 0.51 0.67        30 pswidpost 0.22 0.15 -0.02 0.28 0.46 0.16 0.31 0.32 0.39 0.46 0.66       31 rhymepost 0.25 0.17 0.21 0.07 0.27 0.24 0.09 0.16 0.26 0.16 0.20 0.17      32 ocpost 0.16 0.20 0.16 0.64 0.10 0.06 0.20 0.03 0.71 0.51 0.39 0.28 0.17     33 intdepost 0.23 0.26 0.05 0.08 0.68 0.29 0.07 0.14 0.31 0.37 0.47 0.37 0.35 0.17    34 findepost 0.20 0.26 0.22 0.24 0.26 0.26 0.30 0.22 0.33 0.33 0.48 0.32 0.28 0.35 0.40   35 intidepost -0.03 0.03 0.13 -0.02 0.16 0.01 0.40 -0.13 -0.09 -0.01 -0.09 0.02 0.01 0.00 0.15 0.17  36 pypost 0.23 0.24 0.05 0.02 0.27 0.03 0.09 0.08 0.15 0.18 0.19 0.21 0.05 0.16 0.29 0.19 0.03 Note. Pypre, Pinyin spelling pretest; Chrpre, Chinese character identification pretest; Vocpre, vocabulary (picture naming) pretest; Lnpre, letter naming pretest; Widpre, word identification pretest; Pswidpre, non word reading pretest; Rhymepre, rhyme awareness pretest; Ocpre, oral cloze pretest; intdepre, initial phoneme deletion pretest; findepre, final phoneme deletion pretest; intidepre, initial phoneme identification pretest; vocpost, vocabulary posttest; Lnpost, letter naming posttest; widpost, word identification posttest; pswidpost, non word reading posttest; rhymepost, rhyme awareness posttest; Ocpost, oral cloze posttest;  intdepost, initial phoneme deletion posttest; findepost, final deletion posttest; intidepost, initial phoneme identification posttest; Sppost, spelling posttest; chrpost, Chinese character identification posttest; pypost, Pinyin spelling, posttest;  widDl, word identification delayed test; lsDl, letter sound knowledge delayed test; pswidDl, non word reading delayed test; spDl, spelling delayed test. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0362561/manifest

Comment

Related Items