Writing Strategies and Strategy-Based Instruction in Singapore Primary Schools [1 ed.] 9781443889377, 9781443885959

Citation preview

Writing Strategies and Strategy-Based Instruction in Singapore Primary Schools

Writing Strategies and Strategy-Based Instruction in Singapore Primary Schools By

Barry Bai

Writing Strategies and Strategy-Based Instruction in Singapore Primary Schools By Barry Bai This book first published 2016 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2016 by Barry Bai All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-8595-9 ISBN (13): 978-1-4438-8595-9

TABLE OF CONTENTS

Acknowledgments ..................................................................................... vii List of Tables ............................................................................................. ix List of Figures ............................................................................................. x Preface ........................................................................................................ xi Chapter One ................................................................................................. 1 Introduction 1.1 Expertise studies and the origin of research on language learning strategies (LLS) 1.2 English as an L2 in the Singapore context 1.3 Approaches to teaching writing in an L2 1.4 A multi-dimensional theoretical framework for literacy and LLS 1.5 The research problem and research questions 1.6 Objectives and significance of the study 1.7 An outline of the book Chapter Two .............................................................................................. 10 Literature Review of Research on Language Learning Strategies 2.1 Definitions and categorization of LLS 2.2 Descriptive LLS studies 2.3 Intervention research on LLS 2.4 Conclusion Chapter Three ............................................................................................ 63 Theoretical Framework for the Present Study 3.1 Cognitive views on language learning 3.2 Socio-cultural views on learning 3.3 Multi-dimensional views on language learning and literacy development 3.4 Conclusion

vi

Table of Contents

Chapter Four .............................................................................................. 95 Research Methodology 4.1 Establishing patterns of strategy use (Phase One) 4.2 Determining the effectiveness of strategy-based instruction (Phase Two) 4.3 Ethical considerations Chapter Five ............................................................................................ 130 Findings and Discussion of Phase One 5.1 Descriptive statistics of overall strategy use 5.2 Relationship between strategy use and English learning outcomes 5.3 Patterns of reported strategy use across the participating schools 5.4 Conclusion Chapter Six .............................................................................................. 155 Findings and Discussion of Phase Two 6.1 The quantitative results 6.2 The qualitative results 6.3 Conclusion Chapter Seven.......................................................................................... 178 Conclusions and Implications 7.1 Summary of major findings 7.2 Theoretical implications 7.3 Pedagogical implications 7.4 Limitations of the study 7.5 Recommendations for future research 7.6 Conclusion References .............................................................................................. 195 Appendices .............................................................................................. 214 Appendix A: Teacher’s Journal Log Appendix B: Researcher’s Observation Record Log

LIST OF TABLES

Table 2.1: LLS Definitions ........................................................................ 11 Table 2.2: Characteristics of Good Language Learners............................. 16 Table 2.3: Learning Strategies by Young Learners ................................... 16 Table 2.4: Strategy Inventory of O’Malley and Chamot ........................... 19 Table 2.5: Oxford’s General Strategy Classification................................. 22 Table 4.1: Writing Strategy Questionnaire (Version Five) Item Analysis ............................................................................................... 98 Table 4.2: SBI Writing Classes and Participants ..................................... 106 Table 4.3: A Comparison of Four Strategy Instruction Models .............. 108 Table 4.4: List of SBI lesson plans .......................................................... 114 Table 4.5: Internal Consistency Reliabilities of the Pre- and Postintervention Questionnaires .............................................................. 123 Table 5.1: Descriptive Statistics of Metacognitive Strategies ................. 131 Table 5.2: Descriptive Statistics of Cognitive Strategies......................... 135 Table 5.3: Descriptive Statistics of Social/Affective Strategies .............. 138 Table 5.4: ANOVA Results of School 1 ................................................. 141 Table 5.5: Descriptive Statistics of Strategy Use by Proficiency of School 1 ......................................................................................... 142 Table 5.6: ANOVA Results of School 2 ................................................. 143 Table 5.7: Descriptive Statistics of Strategy Use by Proficiency of School 2 ......................................................................................... 144 Table 5.8: ANOVA Results of School 3 ................................................. 145 Table 5.9: Descriptive Statistics of Strategy Use by Proficiency of School 3 ......................................................................................... 146 Table 5.10: ANOVA Results of School 4................................................ 147 Table 5.11: Descriptive Statistics of Strategy Use by Proficiency of School 4 ......................................................................................... 148 Table 5.12: ANOVA Results of School 5................................................ 150 Table 5.13: Descriptive Statistics of Strategy Use by Proficiency of School 5 ......................................................................................... 151 Table 6.1: Descriptive Statistics of Pre-, Post-, and Delayed Tests ......... 156 Table 6.2: T Test on Frequency of Strategy Use Comparing Experimental and Control Groups ..................................................... 157 Table 6.3: Descriptive Statistics of Self-initiation Strategies .................. 158 Table 6.4: Descriptive Statistics of Monitoring and Evaluating .............. 159

viii

List of Tables

Table 6.5: Descriptive Statistics of Resourcing Strategies ...................... 160 Table 6.6: Descriptive Statistics of Social/Affective Strategies .............. 160 Table 6.7: Descriptive Statistics of Planning Strategies .......................... 161 Table 6.8: Descriptive Statistics of Feedback Handling Strategies ......... 162 Table 6.9: Descriptive Statistics of Revising Strategies .......................... 162 Table 6.10: Descriptive Statistics of Text-generating Strategies ............. 163

LIST OF FIGURES

Figure 3.1: A general framework of Anderson’s ACT production system ................................................................................................. 68 Figure 3.2: Dimensions of literacy ............................................................ 88 Figure 3.3: Summary of dimensions of literacy ........................................ 93 Figure 4.1: A strategy-based instruction model ....................................... 110 Figure 7.1: Dimensions of language learning strategies .......................... 183

ACKNOWLEDGEMENTS

For this book, I would like to thank Dr. Guangwei Hu from the National Institute of Education, Nanyang Technological University, Singapore for inspiring me to conduct my doctoral research. His knowledge has kept me on the right research trajectory, and his insights have compelled me to work hard and write this book. I would also like to express my gratitude to Dr. Peter Gu from the University of Victoria, Wellington, who took me on the journey of research and encouraged me to pursue my studies. Special thanks go to Dr. Lawrence Zhang from the University of Auckland, who has been my wonderful friend and critical reader. I am also grateful to Professor Sandra Mckay from San Francisco State University for her advice and encouragement. I would like to thank the Centre for Research in Pedagogy and Practice, the National Institute of Education, Nanyang Technological University, Singapore for providing me with the opportunity to participate in the “English Language Learning Strategies in Singapore Primary Schools” (CRP 3/03 GYQ) project, from which the data for this book were drawn. Many thanks to my friends and mentors who offered their help, advice, and suggestions. Last but not least, my dearest family members have given me unwavering support. To them, I am greatly indebted.

PREFACE

I am pleased to share this work with you. The study is based on my doctoral research. It provides a theoretical and practical understanding of the writing strategies used by Singapore primary school pupils and the strategy-based writing instruction provided in Singapore primary schools. Research on language learning strategies (LLS) has been an area of inquiry for more than thirty years now. The majority of LLS studies are descriptive studies, many of which have found significant correlations between strategy use and language proficiency. However, the intervention studies, limited in quantity, have yielded inconsistent findings. To further complicate the issue, many of the LLS studies were conducted from a singular theoretical perspective of cognitive psychology. Therefore, these studies have not employed diverse theoretical lenses to fully capture the complexity and plurality involved in language learning. The present study addresses this gap in research through investigating the relationship between Singapore primary school pupils’ use of writing strategies and their writing competence in English. In addition, the study generates empirical evidence about the effectiveness of strategy-based instruction (SBI). Specifically, this two-phase study aims to answer the these questions: what writing strategies Singapore primary school pupils use in order to develop their writing competence in English; what the differences between academically successful primary school pupils and their less successful counterparts are in their use of writing strategies; How the differences in strategy use are related to language competence; and whether strategy-based writing instruction helps pupils improve their writing competence in English and the use of the target writing strategies. The study was carried out within an overriding theoretical framework built on Kucer’s multi-dimensional view of literacy, which incorporates different dimensions, i.e., cognitive, linguistic, socio-cultural, and development dimensions. Such a theoretical framework allows for examination of LLS across different dimensions, thus addressing the multi-dimensional nature of LLS. Consequently, an effective intervention can be designed.

xii

Preface

During Phase 1, planning strategies, text-generating strategies, feedback handling strategies, and revising strategies were found to have a significant positive relationship with English language learning outcomes. The analyses of Phase 2 data show that the SBI had a significant treatment effect on both the participants’ writing competence and their strategy use, namely, text-generating, feedback handling, and revising. The qualitative analyses also show that the experimental pupils orchestrated their strategy use significantly better than before. Taken together, the findings of the study suggest that there are indeed strategic differences between successful pupils and their less successful counterparts in Singapore primary schools, which in turn has led to different English language learning outcomes. Strategy-based writing instruction has produced solid evidence that SBI is an effective method in teaching writing to primary school pupils in Singapore. In light of these findings, both theoretical and pedagogical implications are discussed in the final chapter. Part of the findings in this book has appeared in published journal articles. Some findings of Phase 1 are published in Bai, Hu, and Gu (2014), and some findings of Phase 2 Bai (2015).

CHAPTER ONE INTRODUCTION

Language learning strategies (LLS) are defined as specific actions, behaviors, steps or techniques that learners use to improve their language learning (Oxford, 1999). In particular, writing strategies are the techniques, either mental or behavioral, that writers employ to enhance their writing. The findings of this study may deepen our understanding of the writing strategies employed by Singapore primary school pupils, the relationship between writing strategies and language proficiency, and the effects of strategy-based writing instruction in the Singapore context. In light of the findings, I have also proposed a new conceptual framework to look at strategy-based instruction (SBI), after discussing different dimensions of literacy in relation to writing strategies training. What makes a learner successful in language learning? How do certain learning strategies help learners promote the learning of language? What are these strategies? What is the relationship between LLS and language learning achievements? In other words, we need to know the required information on developing language learners’ learning expertise, such as LLS and successful LLS training (Johnson, 2005). LLS researchers, among researchers from other fields in second language acquisition, are interested to find out whether learning strategies make a difference between more successful learners and less successful ones. More importantly, further investigations on how LLS can be taught effectively to language learners are needed (Griffiths, 2015). In this introductory chapter, I firstly examine the “expertise” research to contextualize LLS research, which stems from the expertise research historically. Next, I discuss the development of second language (L2) writing competence, followed by a multi-dimensional view on literacy as well as SBI. Then, I state the research problem and research questions, present the objectives and significance of the study. Finally, I conclude the chapter with an outline of the book.

2

Chapter One

1.1 Expertise studies and the origin of LLS research The development of expertise research mainly stems from the need to create machines, such as computers, that can think or process like human beings and develop human skills. During the last few decades, expertise research has attempted to identify what accounts as expertise in a number of non-linguistic domains, such as chess playing, problem-solving, mathematics, and even car driving mechanics (e.g., Akin, 1980; De Groot, 1978; Gentner, 1988; Glaser & Chi, 1988; McLaughlin, 1979; Solso & Dallop, 1995, cited in Johnson, 2005). The common characteristics identified across the above domains include, but are not limited to, the following: a) Experts excel mainly in their own domains; b) Experts perceive large meaningful patterns in their domain; c) Experts are fast; they are faster than novices at performing the skills of their domain, and they quickly solve problems with little error; d) Experts have superior short- and long-term memories; e) Experts see and represent a problem in their domain at a deeper (more principled) level than novices; novices tend to represent a problem at a superficial level; f) Experts spend a great deal of time analyzing a problem qualitatively; and experts have strong self-monitoring skills (Johnson, 2005, p. 19-20). There is one assumption that may hold true with regard to the above expertise studies. There may be general learning theories that can be applied to all these various domains, including learning an L2. Indeed, Anderson’s (1982) cognitive theory suggests that learners share common cognitive processes for learning, irrespective of domains. In this view, language learners are perceived as organizers of information, using their information processing capacity to transact with the outside world. Therefore, language learning research can draw on the theories underlying the expertise research. The characteristics of language learning experts and the teaching of these characteristics will benefit other learners. Influenced by the expertise studies, the field of second language acquisition has seen a flourish of studies of the “good language learner” since the 1970s (e.g., Rubin, 1975; Stern 1975). At the outset, Rubin proposed a model that characterized the “good language leaner” with respect to the strategies used by successful L2 learners. The idea was embraced by Stern (1975),

Introduction

3

who identified a series of strategies associated with good language learners. These studies have revealed that good language learners, similar to the experts in other domains mentioned previously, share common characteristics. They are active in learning, monitor their language production, make use of their prior language knowledge, employ memorization techniques, and clarify when there is a problem (Chamot, 2001). These characteristics of language experts share striking similarities with those identified in the expertise research mentioned above. The earliest concerns and research work with the “good language learner” spearheaded the field of LLS. The main focus of the early work was to describe which strategies successful language learners used that poor learners did not so that poor learners could learn from the successful ones. The majority of LLS studies have been devoted to identifying LLS used by language learners. Research has shown that effective language learners use strategies more appropriately than those less effective ones, and that learning strategies can be taught to less effective learners (e.g., Carrier, 2003; Chamot & El-Dinary, 1999; Gu, Hu, & Zhang, 2005; Kuba, 2002; Lan & Oxford, 2003; Magogwe & Oliver, 2007; Nong-Nam & Leavell, 2006; O’Malley & Chamot, 1990; Vandergrift 2003a). Two general strands of LLS studies coexist in the literature: descriptive studies and intervention studies (McDonough, 1995). Most descriptive studies have been devoted to identifying the strategies used by successful language learners, investigating the relationship between learners’ LLS use and their language learning achievements, and examining learners’ strategic performance in different language skill areas. Though the LLS literature presents a rather complex relationship between LLS and language learning achievements, a large repertoire of LLS that are associated with good language results has been identified. In addition, the majority of LLS studies have shown that good language learners are good at matching their use of strategies to the tasks at hands, whereas their weak counterparts lack the metacognitive knowledge about the task requirements for choosing appropriate strategies. On the other hand, a number of LLS researchers have also undertaken studies aimed at training learners to employ LLS. The intervention studies include those that attempt to teach strategies for overcoming problems encountered in different aspects of language learning, and the ones that aim to teach particular strategies, e.g., reading strategies, writing strategies, vocabulary learning strategies, listening strategies, and speaking strategies.

4

Chapter One

These intervention studies are generally termed as strategy-based instruction (SBI). Motivated by this line of research, I am keen to know more on how to help language learners improve their language proficiency. Why do some language learners seem to learn a language very successfully, but some are not successful? I am very interested to investigate what the LLS are that positively correlate with learners’ language achievements and how the training of strategy use can be successful. This study is connected with both strands of LLS studies.

1.2 English as an L2 in the Singapore context Due to the multi-racial composition of the Singaporean population, English serves as a lingua franca among the major ethnic groups (i.e., Chinese, Malay, Tamil, and Eurasian). English plays an essential role in business, technology, administration, and the medium of instruction in school. In Singapore, primary school students study at least two languages, namely, English and their mother tongue. While most Singaporean children are bilingual, with English-cum-Mother Tongue combination, many children speak and interact with their family members and friends in their mother tongue at home. As shown in the Singapore 2010 population census, only 32.29% of Singapore families used English for communication among their family members (Singapore Department of Statistics, 2011). However, English, the medium of instruction for education, is used predominantly in Singapore schools. Given this situation, English is generally perceived as an L2 in the Singapore context by many researchers and teachers.

1.3 Approaches to teaching writing in an L2 It is also useful to look at the development of writing instruction in an L2 since the present study is concerned with writing strategies and strategybased writing instruction. Product, current-traditional rhetoric, and process approaches are the three well-established traditions for teaching writing. There are no absolutely distinguishable boundaries that make the three approaches completely distinct from each other in the classroom. These approaches may be used by the same teacher at different times for writing instruction. Product-based approaches focus primarily on formal accuracy and correctness of students’ writing. Instructions following these approaches prefer grammar study, error analysis, sentence combination, and focus on the text itself (Silva 1990). On the other hand, the currenttraditional rhetoric approaches (also known as genre-based in some

Introduction

5

contexts) to teaching writing aim to orient the writers with the characteristically patterned ways of writing. The focus is to construct and arrange the discourse forms logically (Silva, 1990). In a typical situation, student writers are instructed to pay attention to topic sentences, support sentences, concluding sentences, and transitions. In addition, they have to arrange sentences and paragraphs into prescribed patterns, e.g., narration, exposition, and arguments. This approach has been criticized due to its prescriptivism. In the Singapore context, these patterns are known as text types and have gained popularity in the writing class. Process-based approaches reflect the trend that attention has shifted from texts to writers. Research evidence suggests that product-based approaches to writing, e.g., grammar instruction, did not directly improve students’ writing ability, which gave rise to researchers’ exploration of writing processes. Writing is not only about recording down thoughts, ideas and sentences, but also about thinking processes, and generating new thoughts and ideas. Moffett (1968, cited in Kern, 2000) notes that learning to write is learning to think about increasingly abstract topics and to think about the communicative needs of increasingly broad audiences. Also different from current-traditional rhetoric approaches by which learners learn how to write according to the model texts, process-based approaches view writers’ processes as essential for learners to learn. Typically, processbased writing rests on procedures for solving problems, exploring ideas, expressing these ideas in writing, and multiple revisions of text, independent of the socio-cultural contexts (Ferris & Hedgcock, 2005). While the current-traditional rhetoric approaches represent a theoretical view of writing as a linear process, process-based approaches perceive writing as a series of recursive processes. Many studies have investigated the writing processes of L2 writers (e.g., Baker, 2004; Chamot & ElDinary, 1999; Leki, 1992; Raimes, 1985; Victori, 1999; Zamel, 1983). The research of writing strategies represents part of the process writing movement (ManchÓn, 2001). This line of research has generated rich insights into what the writing strategies are that lead to successful writing. Despite the importance accorded to writing, many primary school pupils still do not do well to meet their grade-level demands and need substantial help in writing (see Graham, MacArthur, & Fitzgerald, 2007). Graham and Perin (2007) attributed this fact to the lack of effective instructional practices for teaching of writing to students. Students who do not learn to write well are at a disadvantage both in school and at the workplace in the future as writing is an extremely important skill for both academic and professional development. In the Singapore context, many young writers

6

Chapter One

are also faced with a similar problem (Goh, Zhang, Ng, & Koh, 2005). A research study on what writing strategies can help Singapore primary school pupils write more effectively will help inform writing teachers with regard to their classroom instructions.

1.4 A multi-dimensional theoretical framework for literacy and LLS As mentioned previously, the LLS research in general and writing strategies research in particular represent a strong cognitive orientation. However, the teaching and learning of writing do not only involve the teaching and learning of a set of cognitive processes, such as inscribing and decoding words or patterned ways of thinking (Kern, 2000). Sociocultural advocates (e.g., Gee, 1996; Vygotsky, 1986) have argued that writing is also a socially rooted process. In their view, learners do not only engage in abstract processes but also interact with others. It is the social interactions that help them develop their writing abilities in particular and literacy in general. The most recent writing strategies research has also seen many studies that have taken a socio-cultural perspective (e.g., Coyle, 2007; Donato & McCormick, 1994; Guerrero & Villamil, 2000; Hu, 2005; Min, 2006). The two aforementioned theoretical perspectives (cognitive & sociocultural) suggest that various disciplines tend to theorize learning in a unitary manner. For example, cognitive psychologists (e.g., Anderson, 1982) stress the importance of mental processes students go through during learning. Socio-culturists (e.g., Gee, 1996, Vygotsky, 1986) emphasize the impacts of social interactions on learners. However, Kucer (2005, 2009) argues each discipline only offers a partial view on literacy. Accordingly, the pedagogical implications as a result of such a narrow view will only lead to ineffective instruction, e.g., exclusive focus on the teaching of writing processes (cognitive). Therefore, Kucer (2005, 2009) has proposed a theoretical perspective that incorporates different dimensions: cognitive, linguistic, socio-cultural, and developmental. For literacy learning, the learner must learn to effectively, efficiently, and simultaneously gain control over the different dimensions of the written language. A learner takes on multiple roles: the role of code breaker/maker (linguistic dimension), the role of meaning making (cognitive dimension), the role of text user and critic (socio-cultural dimension), and the role of scientist and construction worker (developmental dimension). Writing instruction, theoretically grounded in a multi-dimensional framework, may

Introduction

7

offer a better potential for teaching writing in the classroom effectively as this framework looks at the writer in a comprehensive manner. In this view, the development of the writer’s competence is examined cognitively, linguistically, socio-culturally and developmentally. In a similar vein, the research of writing strategies should also be explored multi-dimensionally, given the importance of writing in the development of young pupils’ literacy.

1.5 The research problem and research questions The previous thirty years of research on LLS has seen a plethora of empirical studies (Cohen & Macaro, 2007). The majority of LLS studies fall into the category of descriptive studies, many of which have found significant correlations between strategy use and language proficiency. However, the intervention studies, limited in quantity, have yielded inconsistent findings. To further complicate the issue, many LLS studies were conducted based on a singular theoretical perspective, i.e., cognitive psychology, failing to address the multi-dimensional nature of LLS theoretically. Therefore, Gu (2007) notes in his Foreword to Cohen and Macaro (2007) that the research need to demonstrate “more rigorous designs and practices, and tangible and useful applications for learners and teachers” is still compelling. Despite the abundant LLS research studies conducted worldwide, surprisingly very few studies (e.g., Goh, 2002; Lee, 2009; Loh, 2007; Wharton, 2000; Zhang, 2008) on language learning strategies employed by Singaporean students have been undertaken. In addition, most of these LLS studies were concerned with LLS used by older learners, e.g., preuniversity or university students. Therefore, there is a need to investigate the LLS used by younger learners, i.e., primary school pupils. What are the LLS used by effective language learners in Singapore primary schools? What is the relationship between their strategy use and language learning achievements? What makes the training of strategy use effective in the Singapore context? In general, pupils who struggle with writing are found to lack knowledge of writing strategies, have difficulties in generating ideas, do little or no pre-writing planning, use few writing strategies, and revise little (Harris, Graham, Mason, & Friedlander, 2008). The problems that hinder them from developing their writing competence should be identified and solved as early as possible. All the above-mentioned questions can be formulated into specific research questions for the present study as stated below.

Chapter One

8

1. 2. 3. 4.

What writing strategies do upper primary school pupils use? How are the differences in strategy use related to language competence? In what ways does strategy-based writing instruction help pupils improve their writing competence in English? In what ways does such instruction help improve pupils’ use of the target writing strategies?

1.6 Objectives and significance of the study There are two phases in this study, with each phase covering one general objective. The first phase is to map out the relationship between writing strategies and language proficiency. In the second phase, the effectiveness of writing strategy intervention with primary school pupils will be examined. As far as the research significance is concerned, the study can offer insights in the following three aspects. First, the study will shed light on what writing strategies primary school students use in the Singapore context so that teachers can conduct suitable training programs for their students to enhance their writing competence. Primary school teachers can also adjust their teaching of writing in accordance with the findings from this study. Second, policy makers and curriculum designers can use the findings as guidelines for designing the writing curriculum. Such a study can provide empirical data for policy makers and curriculum designers to better shape the teaching of writing in Singapore primary schools. Third, the research findings are expected to contribute particularly to LLS research in Singapore and similar educational settings worldwide.

1.7 An outline of the book This book comprises seven chapters. This chapter has briefly introduced the research field of LLS in relation to expertise research, development of language learning strategies research, and a multi-dimensional view on literacy as well as LLS, thus providing a broad context for the present study. In chapter 2, I will examine the development of LLS research by reviewing the LLS literature. Such themes as definitions and classifications of LLS, descriptive studies, and intervention studies will be reviewed. Chapter 3 will discuss the theoretical framework on which the present study is based. Both cognitive and socio-cultural theories will be drawn upon to inform the present study theoretically. In addition, a multi-

Introduction

9

dimensional view on LLS and SBI will also be reviewed to gain a better theoretical understanding of both LLS and SBI. Chapter 4 will be devoted to presenting the research methods of the study. As mentioned previously, the study consists of two phases. The research methods associated with each phase will be presented separately. In chapter 5, I will present the research results and discussion concerning Phase 1. Chapter 6 will present and discuss the research findings of Phase 2. Finally, chapter 7 will summarize the major findings of the two phases briefly. A discussion of both theoretical and pedagogical implications in accordance with the research findings will be provided. In addition, limitations and suggestions for future research will also be discussed.

CHAPTER TWO LITERATURE REVIEW OF RESEARCH ON LANGUAGE LEARNING STRATEGIES

This chapter opens with a detailed review of the frequently cited definitions and classifications in the literature for the purpose of defining and classifying LLS for the present study. Then, it continues to examine descriptive studies of LLS. A review of this body of research will show the commonly found relationships between strategy use and language learning outcomes. I will also critique some of the issues with respect to descriptive studies. Finally, the chapter concludes with a review of empirical studies that involved LLS training. The empirical findings of these studies can inform the present study on the intervention design that is most likely to be effective.

2.1 Definitions and categorizations of LLS Despite the substantial LLS research thus far, LLS researchers not only tend to define learning strategies differently, but also categorize them differently, which brings about confusion to the research field. In section 2.1.1, I will focus on reviewing the definitions in the literature. In section 2.1.2, I will look at the issues on LLS classifications.

2.1.1 Definitions It is important to note that learning strategies must not be confused with communication strategies although there is some overlap. Whereas learning strategies are used to facilitate learning, communication strategies are employed in order to overcome communicative problems. This line of research will not be covered in this review. Although LLS research has been underway for more than three decades, disagreement exists among researchers with respect to the definition of LLS. It seems that the majority of LLS research has been interested in more practical goals, i.e., to explore ways of empowering language learners to be more effective in learning,

Literature Review of Research on Language Learning Strategies

11

rather than “pushing the theoretical understanding of language learning strategies forward” (Tseng, Dörnyei, & Schmitt, 2006, p. 78), thus calling for more research into the theoretical understanding of LLS. In the LLS literature, there exist varying definitions (see Table 2.1 for some of them). Among them, Oxford’s (1989, 1999) and O’Malley & Chamot’s (1990) definitions have been widely used by LLS researchers with some adaptation to conceptualize their own studies. To these researchers, language learning strategies are special techniques, ways, approaches, or behaviors to facilitate information processing so that language learning can be enhanced. Table 2.1. LLS Definitions Source Stern (1983, p. 405)

Weinstein & Mayer (1986, p. 315) Chamot (1987, p. 71)

Rubin (1987, p. 23) Wenden 1987a (1987a, p. 6) Oxford (1989, p. 235) Oxford (1990, p. 1) O’Malley & Chamot (1990, p. 1)

Definition “In our view strategy is best reserved for general tendencies or overall characteristics of the approach employed by the language learner, leaving the techniques as the term to refer to particular forms of observable learning behavior.” “Learning strategies are the behaviors and thoughts that a learner engages in during learning that are intended to influence the learner’ encoding process.” “Learning strategies are techniques, approaches or deliberate actions that students take in order to facilitate the learning, recall of both linguistic and content area information.” “Learning strategies are strategies which contribute to the development of the language system which the learner constructs and affect learning directly.” “…the term learner strategies refers to language learning behaviors learners actually engage in to learn and regulate the learning of a second language.” “Language learning strategies are behaviors or actions which learners use to make language learning more successful, self-directed and enjoyable.” “Language learning strategies are steps taken by students to enhance their own learning.” “Special thoughts or behaviors that individuals use to help them comprehend, learn, or retain new information.”

12

Cohen (1998, p.5)

Oxford (1999, p. 518)

Weinstein, Husman, & Dierking (2000, p. 727)

Chapter Two

“Language learning strategies include strategies for identifying the material that needs to be learned, distinguishing it from other material if need be, grouping it for easier learning (e.g., grouping vocabulary by category into nouns, verbs, adjectives, adverbs, and so forth), having repeated contact with the material (e.g., through classroom tasks or the completion of homework assignments), and formally committing the material to memory when it does not seem to be acquired naturally (whether through rote memory techniques such as repetition, the use of mnemonics, or some other memory technique).” “Specific actions, behaviors, steps, or techniques that students use to improve their own progress in developing skills in a second or foreign language. These strategies can facilitate the internalization, storage, retrieval, or use of the new language.” “Learning strategies include any thoughts, behaviors, beliefs, or emotions that facilitate the acquisition, understanding, or later transfer of new knowledge and skills.”

As shown in the definitions above, LLS researchers (e.g., Chamot, 1987; Oxford, 1989; Rubin, 1987; Wenden, 1987a; Weinstein, Husman, & Dierking, 2000; Weinstein & Mayer, 1986) could not agree on what counts as a strategy. According to Oxford (1989, p. 235), “Language learning strategies are behaviors or actions which learners use to make language learning more successful, self-directed and enjoyable.” By this definition, learning strategies are essentially behavioral and can be observed. Weinstein and Mayer (1986), Chamot (1987), and O’Malley and Chamot (1990), however, regard LLS as both behavioral and mental. Their definitions indicate that some strategies cannot be observed directly since they are mental. For example, translation (i.e., using the learner’s first language to understand or produce the second language) is such a strategy. The disagreement on whether strategies are observable seems to be resolved by Oxford’s (1990) definition of strategies as “steps taken by the learner” rather than “behaviors or actions which learners use.” “Steps” can be mental processes as well as behavioral ones. Improving on her previous attempts in defining a strategy, Oxford (1999) offers a more inclusive definition. According to this definition, strategies are “specific actions, behaviors, steps, or techniques.” Compared to the definitions above, which

Literature Review of Research on Language Learning Strategies

13

are somewhat abstract, Cohen (1998) provides concrete examples in his definition (see Table 2.1 for the examples). However, whether strategies are mental or behavioral are still debatable. The disagreement seems to have been resolved by Chamot (2004), who notes that “Learning strategies are for the most part unobservable, though some may be associated with an observable behavior” (p.15). As a military term, “strategy” has been used to refer to procedures for implementing the plan of a large-scale military operation. This term, when used in other contexts, refers to procedures for accomplishing something, e.g., language learning (Schmeck, 1988). However, confusion about the difference among “strategies”, “techniques”, “actions”, “steps”, and other possible terms exits among LLS researchers. Ellis (2008) attributes the confusion to the uncertain nature of the behaviors that are considered as learning strategies. Stern (1983) draws a distinction between the two terms, i.e., strategies and techniques. Strategies are seen as general or overall approaches used by the language learner (e.g., an active task approach), whereas techniques refer to particular forms of actions that can be observed (e.g., using a dictionary when necessary). However, other researchers do not make such a distinction. They generally treat Stern’s “techniques” as strategies too, for example, as can be seen in Rubin’s (1987) definition. Cohen (1998), after a thorough review of the terms (e.g., strategy, macro-strategy, micro-strategy, technique, tactic, and move) used in the literature to refer to various cognitive or metacognitive processes, proposes a solution to this problem. He proposes to refer to all of the existing terms as “strategy”. Among LLS researchers, whether learning strategies are deployed consciously and intentionally or subconsciously by the language learner is also a debatable issue. Chamot (1987) perceives learning strategies as conscious and intentional, by referring to them as “deliberate actions”. Cohen (1998) views consciousness as a crucial factor which distinguishes strategies from those processes that are not strategic. The same view is echoed by Anderson (2005) as well. However, other researchers (e.g., O’Malley & Chamot, 1990; Oxford, 1989) leave this factor out in their definitions. Dörnyei and Skehan (2003), in an attempt to differentiate strategies from skills, offer their view towards the consciousness argument on strategies. They mention that “strategies and tactics involve the conscious decisions …” (p. 611). Consciousness is surely an issue in question in defining strategies. In attempting to resolve this disagreement, Gu (2012) puts forward a new perspective, a prototypical view, on the

14

Chapter Two

nature of learning strategies. In his view, some strategies are more strategy-like than others, varying in degree. Strategy is not a static concept but a dynamic process in nature. Therefore, some strategies might involve more consciousness, and some less. He explains: Prototypes are the ideal forms, so to speak, of target concepts. Particular instances are evaluated by means of comparing them to the prototypical exemplars to see how much common variance they share. Finding strategy prototypes and matching various strategic properties against them offers a much more illuminating perspective in the definition and description of strategies than simple categorizations based upon the presence or absence of, for instance, generality, or of other strategic attributes. Hence, the definition of learning strategies, according to prototype theory, would mean the delimitation of attributes that anchor the central core of a strategy, while at the same time spelling out possible dimensions of variance in much the same way as, though far more complicated than, Labov’s (1973) demonstration of the concept of a cup. (p. 336)

Similar to Gu (2012), Afflerbach, Pearson, and Paris (2008) seem to have resolved the issue by proposing that learner’s deliberate control, goal directedness, and awareness characterize learning strategies. However, learning skills are used without learners’ deliberate control or consciousness and they are normally used out of habit and automatically. The key difference between skill and strategy lies in their intentionality and automaticity. When a strategy, which initially requires learners’ deliberate attention, is used with no effort and automatically by learners, it has become a skill. Conflicting views also exist as to whether the effect of learning strategies on language learning is direct or indirect in the definitions. Rubin’s (1987) definition asserts that learning strategies produce a direct effect on language learning. By her definition, learning strategies contribute to the development of the evolving language system directly. But other researchers hold that the effect is more indirect. For example, Weinstein, Husman, and Dierking (2000) emphasize that strategies are to facilitate the learning process, similar to Oxford’s (1999) definition. LLS researchers also have different perceptions on the motivation for using learning strategies. Oxford, in her 1989 definition, considers that making learning an L2 “enjoyable” constitutes the motivation for strategy use as well. However, other researchers (e.g., O’Malley & Chamot, 1990; Rubin, 1987; Weinstein et al., 2000) believe that learning something about the L2 motivates the learner to use language learning strategies, excluding

Literature Review of Research on Language Learning Strategies

15

the “enjoyment” element in their definitions. It is important to point out that enjoyment is only a minor motivation. By definition, the major motivation is to solve a language learning problem or to improve language learning.

2.1.2 Classifications As it is with definitions of strategy, classifications of learning strategies also vary due to researchers’ different understandings of learning strategies. Early researchers (e.g., Naiman, Fröhlich, Stern, & Todesco, 1978; Rubin, 1975, 1981; Stern, 1975; Wong-Fillmore, 1976) focused on compiling inventories of LLS that the learners were observed to use or reported using. These pioneer studies made no attempt to classify them into specific categories. Subsequent research, for example, by O'Malley, Chamot, Stewner-Manzanares, Küpper, and Russo (1985a, 1985b), O’Malley and Chamot (1990), Wenden (1991) and Oxford (1990), built on those early studies to develop their individual classification schemes. In this section, I will first review the inventories that the pioneer researchers compiled and then examine the classifications developed in subsequent research. Rubin (1975) listed a series of characteristics that good language learners possess. According to Rubin, good language learners are willing to guess and can guess accurately, have a strong drive to communicate, are often uninhibited, focus on communication and attend to form, practice and monitor their own speech and the speech of others, and attend to meaning. Rubin (1981) conducted a longitudinal study eliciting and classifying language learning strategies of young adult learners. She used directed diary-reporting, focusing on particular types of cognitive strategies, instead of the whole range of strategies. Two types of strategies were identified: strategies which may contribute directly to learning, and those that contribute indirectly (see Table 2.2). As Rubin used basic psychological processes (e.g., monitoring, memorization, and guessing) to describe learning strategies found in the learners, both of the two categories she put forward bear a cognitive orientation. Rubin’s (1981) inventory contained mainly cognitive strategies. Metacognitive and social/affective strategies were not included.

16

Chapter Two

Table 2.2. Characteristics of Good Language Learners Direct strategies 1. clarification/verification 2. monitoring 3. memorization 4. guessing/inductive inferencing 5. deductive reasoning 6. practice

Indirect strategies 1. creation of opportunities for practice 2. production tricks (e.g., using formulaic interaction, circumlocution, synonyms or cognates)

In a similar vein, Stern (1975) proposed a list of good language learners’ learning behaviors and styles based on his survey of the research literature and his experience as a language teacher and learner. The 10 items were not categorized into any groups. It appears that Stern’s inventory just simply listed the general good learning behaviors and styles exhibited by the learners. For example, the first item in the list is actually a “personal learning style” (p. 316). Wong-Fillmore (1976, 1979) identified three social and five cognitive strategies (see Table 2.3) in her study on five children’s language learning strategies over a nine-month period. Wong-Fillmore’s interpretation is that the three social strategies were more important than the five cognitive strategies as the children were more concerned with establishing social relationships with each other when learning language. English was the tool used to set up relationships between the five Mexican children and their American peers, so they had to learn it. Table 2.3. Learning Strategies by Young Learners Social strategies 1. Join a group 2. Give the impression you can speak the language 3. Count on your friends

Cognitive strategies 1. Assume relevance of what is being said 2. Get some expressions 3. Look for recurring parts in formulae 4. Make the most of what you’ve got 5. Work on the big things first

Literature Review of Research on Language Learning Strategies

17

According to her account, the first two social strategies were well linked to the first two cognitive strategies, respectively. The children had to employ the first social strategy as an entry token to group membership. After successful entry, they used the first cognitive strategy of assuming they were in the relevant context. Then, they had to justify their membership by using the second social strategy, i.e., “give the impression, with a few well-chosen words, that you speak the language” (Wong-Fillmore, 1979, p. 209). The accompanying cognitive strategy was to “get some expressions you understand, and start talking”. As shown in the above review, WongFillmore’s inventory grounded LLS mainly in the domain of communication, participants’ interactions while learning English. Therefore the three social strategies are mainly communication-orientated in nature. Only two categories of strategies, i.e., cognitive and social strategies, were included in this inventory. Metacognitive strategies, a very important category, were left out. However, the study is still of significance to the present study of LLS because Wong-Fillmore pointed out the importance of orchestration of strategies. For example, she noted that the first two social and cognitive strategies were orchestrated by the children in learning English. Inspired by Stern, Naiman et al. (1978) conducted a “good language learner” study which involved 34 successful adult language learners and a group of secondary school French-as-L2 learners. By means of interviews and observations, the researchers identified five broad groups of strategies. They discovered that good language learners take an active approach to their language learning, are able to develop an awareness of language both as a system of rules and as a means of communication, know they need to manage affective demands of language learning, and monitor their L2 production. These five broad groups each contain a group of secondary strategies. For example, awareness of language both as a system of rules and as a means of communication comprises emphasizing fluency over accuracy and seeking communication situations with L2 speakers. These strategies were elicited from adult learners and included more strategies as compared with Wong-Fillmore’s (1979) and Stern’s (1975) inventories. The above-mentioned early studies were mainly exploratory in nature. The researchers conducted their research in different settings and with different methodological approaches. Wong-Fillmore observed children’s interactions while learning English, so her inventory of LLS exhibited a communicative nature. Rubin’s (1981) inventory is cognitive-orientated as she was only concerned with the psychological processes of the learners.

18

Chapter Two

Stern (1975), on the basis of a literature review and personal teaching experience, described broad learning behaviors and styles of effective language learners. It is not plausible to make generalizations based on the data collected from the above-mentioned case studies. Therefore, more research is yet to be conducted to refine LLS inventories through various methodological approaches. Skehan (1989) identified three commonalities among those early inventories of learning strategies, which are “the learner’s capacity to impose himself on the learning situation”; “his technical predispositions”; and “his capacity to evaluate” (p. 81). These early studies of LLS were mainly focused on identifying the LLS used by the language learners, whereas later studies which will be reviewed below not only examined the language learners’ strategy use, but also investigated how strategies could be classified with more refined research methods. Consequently, these inventories have been more wide-ranging and systematic. O'Malley et al. (1985a) study set out to 1) identify the range, type and frequency of learning strategies that beginning and intermediate level ESL students employed, and 2) determine the types of language tasks that the learning strategies were associated with. They used interviews and observations to collect data on the different learning strategies used for various types of language learning activities by the ESL learners in a secondary school setting, and classified the twenty-six strategies they found into three categories: metacognitive strategies, cognitive strategies, and social mediation. On the basis of the above scheme, O’Malley and Chamot (1990) proposed a slightly different taxonomy. It is composed of three categories: metacognitive strategies, cognitive strategies, and social/affective strategies (see Table 2.4). Metacognitive strategies are “higher order executive skills that may entail planning for, monitoring or evaluating the success of a learning activity” (O’Malley & Chamot, 1990, p. 44). Strategies, such as directed attention and self-evaluation, are subsumed in this category, and they have an executive function as they involve thinking about learning processes. Learners without metacognitive strategies are learners without directions. Therefore, importance is given to metacognitive strategies. Cognitive strategies “operate directly on incoming information, manipulating it in ways that enhance learning” (O’Malley & Chamot, 1990, p. 44). Examples of cognitive strategies are inferencing, or guessing from the context, and elaboration, or relating new information to other concepts in memory. Hence, cognitive strategies have an operative or cognitive-

Literature Review of Research on Language Learning Strategies

19

processing function. Social/affective strategies represent “a broad grouping that involves either another person or ideational control over affect” (O’Malley & Chamot, 1990, p. 45). This category of strategies includes such strategies as cooperation, questioning for clarification, and self-talk. Table 2.4. Strategy Inventory of O’Malley and Chamot Strategy classification Metacognitive strategies

Representative strategies Selective attention

Planning Monitoring

Evaluation

Cognitive strategies

Rehearsal Organization

Inferencing Summarizing

Deducing

Definitions Focusing on special aspect of learning tasks, as in planning to listen for key words or phrases. Planning for the organization of either written or spoken discourse. Reviewing attention to a task, comprehension of information that should be remembered, or production while it is occurring. Checking comprehension after completion of a receptive language activity, or evaluating language production after it has taken place. Repeating the names of items or objects to be remembered. Grouping and classifying words, terminology, or concepts according to their semantic or syntactic attributes. Using information in text to guess meanings of new linguistic items, predict outcomes, or complete missing parts.

Chapter Two

20

Imagery

Transfer Elaboration

Social/affective strategies

Cooperation

Questioning for clarification Self-talk

Intermittently synthesizing what one has heard to ensure the information has been retained. Applying rules to the understanding of language. Using visual images (either generated or actual) to understand and remember new verbal information. Using known linguistic information to facilitate a new learning task. Linking ideas contained in new information, or integrating new ideas with known information. Working with peers to solve a problem, pool information, check notes, or get feedback on a learning activity. Eliciting from a teacher or peer additional information, rephrasing, or examples. Using mental redirection of thinking to assure oneself that a learning activity will be successful or to reduce anxiety about a task.

A few features of O’Malley and Chamot’s classification (1990) are worth noting. First, many of the strategies identified by Naiman et al. (1978) and Wong-Fillmore (1976; 1979) are not found here. For example, the strategy group, i.e., realization of language as a means of communication and management of affective demands (Naiman et al., 1978), is not included in O’Malley and Chamot’s classification. No strategies that are compatible with Wong-Fillmore’s social strategies are identified or reported in their classification. Also, there is little emphasis on the interaction and input maximizing strategies that Naiman et al. (1978) and Wong-Fillmore were so concerned with. The differences may reflect whether the setting is a

Literature Review of Research on Language Learning Strategies

21

formal or informal one. It is likely that the strategies used by learners in the classroom and those in a natural environment are different. For example, Wong-Fillmore’s young learners’ primary needs were to be accepted by their playmates, whereas O’Malley et al.’s (1985a) secondary school students were more concerned with their own learning and achievements in school. In addition, the difference might arise from the fact that O’Malley and Chamot’s classification (1990) was developed mainly based on cognitive theories. Another difference between O’Malley and Chamot’s classification and the early inventories (Naiman et al., 1978; Rubin, 1975, 1981; Wong-Fillmore, 1976, 1979) is the greater focus on metacognitive strategies. O’Malley and Chamot identified nine strategies in the metacognitive category. However, Naiman et al. (1978) and Rubin (1981) identified only one metacognitive strategy, i.e., monitoring, in their respective inventories. The role of metacognitive strategies was not spelled out by these researchers. To O’Malley et al. (1985a, p, 24), “Students without metacognitive approaches are essentially learners without directions and ability to review their progress, accomplishment, and future learning directions”. This basically reflects the researchers’ concern for the role of learners as selfreflective beings, with insight into, and control over their own learning process to maximize learning. Oxford (1990) incorporated all the strategies found in previous research in a single classification scheme. She divides the strategies into two general groups: direct strategies and indirect strategies (see Table 2.5). According to Oxford, direct strategies are learning strategies that directly involve the language being learnt. This group comprises three categories: memory strategies, cognitive strategies, and compensation strategies. The three categories of strategies all require mental processing of the language being learnt, but with different processes and for different purposes. Indirect strategies help facilitate language learning and provide indirect support for language learning. The three categories under this group are metacognitive strategies, affective strategies, and social strategies. The indirect strategies provide “indirect support for language learning through focusing, planning, evaluating, seeking opportunities, controlling anxiety, increasing cooperation and empathy, and other means” (Oxford, 1990, p. 151).

22

Chapter Two

Table 2.5. Oxford’s General Strategy Classification Direct strategies 1. Memory strategies 2. Cognitive strategies 3. Compensation strategies

Indirect strategies 1. Metacognitive strategies 2. Affective strategies 3. Social strategies

Oxford’s classification, though building on some earlier classifications, differs from them in several ways. First, her classification may resemble Rubin’s (1981) on the surface, for both make a distinction between direct and indirect strategies. However, the majority of strategies identified by Rubin tend to be cognitively oriented. As discussed previously, she used basic psychological processes already identified in general learning (e.g., clarification/verification, monitoring, and memorization) to categorize strategies employed by language learners. Oxford, on the other hand, based her classification on a broader view towards the language learner. The strategies of a cognitive orientation only constitute the category of direct strategies. For Oxford, different categories of strategies function for different purposes and different aspects of communicative competence (e.g., sociolinguistic competence, discourse competence, strategic competence). Development of the language learner’s communicative competence, i.e., competence or ability to communicate, is the main goal for using language learning strategies. Cognitive strategies, such as analyzing and particular memory strategies, like the keyword strategy, would be very useful for the learner to become competent in using the new language (e.g., understanding and recalling new information). Affective strategies serve to develop self-confidence and perseverance of the learner. Social strategies, such as cooperation with either native speakers or peers, help develop sociolinguistic competence. Another difference between Oxford’s (1990) classification and Rubin’s (1981) is their different understandings of directness and indirectness. In Rubin’s terms, direct strategies are associated with on-the-spot learning, such as clarification/verification, and guessing/inductive inference. The strategies used for out-of-class activities are indirect strategies. They include monitoring, memorization, and practice. However, Oxford classified all those strategies (i.e., memory, cognitive, and compensation) that require

Literature Review of Research on Language Learning Strategies

23

mental processing of the language as direct strategies, and those (i.e., metacognitive strategies, affective strategies, and social strategies) that support and manage language learning as indirect strategies. There are also differences between Oxford’s (1990) classification and O’Malley and Chamot’s (1990). In Oxford’s scheme, cognitive and memory strategies are two separate categories, whereas the two are grouped as cognitive strategies by O’Malley and Chamot. According to Oxford, imagery, which is subsumed under the category of cognitive strategies in O’Malley and Chamot’s system, is one of the memory strategies, which help the learner to store and retrieve new information. The compensation strategies identified by Oxford, which are normally called communication strategies (Cohen, 1998), are not included in O’Malley and Chamot’s system at all. Despite the difference, Oxford’s (1990) and O’Malley and Chamot’s (1990) systems are quite compatible. The metacognitive strategies category in Oxford’s system is a direct equivalent to the one in O’Malley and Chamot’s. Oxford’s social and affective strategies roughly resemble O’Malley and Chamot’s social/affective strategies.

2.1.3 Critique of definitional and classificatory issues in LLS research Researchers (e.g., Dörnyei, 2005; Ellis, 2008; Macaro, 2006) have identified several problems with regard to the definitions of LLS that the literature presents. As noted by Ellis (2008), there are a number of problems associated with the definitions and classifications of LLS in the literature. Theoretical inconsistencies among different researchers have led to a confusing plethora of LLS definitions. First, the majority of LLS researchers did not ground their research within a coherent theory, which is evident in their varying strategy definitions. Second, although theories of cognitive psychology have been applied in research of language learning strategies (Macaro, 2006), few researchers have consistently situated their research within cognitive theory. Given the lack of definitional agreement, Oxford and Crookall (1989) thus called for consensus on definitions among researchers: The field would be helped considerably if researchers could come to some consensus on definitions of various strategies. As yet there is neither agreement on an overall, hierarchically organized LLS taxonomy or

Chapter Two

24

typology, nor on ways to delimit or define a given strategy or cluster of strategies. Different researchers use different terms and concepts, although in the last few years a little progress has been made in this area. Greater agreement and more standardized procedures will increase comparability across studies, allowing results to have greater generalizability and explanatory power. (p. 414 - 415)

Ellis (2008) proposed that one of the best approaches to defining learning strategies is to identify their main characteristics. A list of the main characteristics was put forward by Ellis: z Strategies refer to both general approaches and specific actions or

techniques used to learn an L2. z Strategies are problem-oriented—the learner deploys a strategy to

overcome some particular learning problem. z Learners are generally aware of the strategies they use and can

z

z z z

z

identify what they consist of if they are asked to pay attention to what they are doing/thinking. Strategies involve linguistic behavior (such as requesting the name of an object) and non-linguistic (such as pointing at an object so as to be told its name). Linguistic strategies can be performed in the L1 and in the L2. Some strategies are behavioral while others are mental. Thus some strategies are directly observable, while others are not. In the main, strategies contribute indirectly to learning by providing learners with data about the L2 which they can then process. However, some strategies may also contribute directly (for example, memorization strategies directed at specific lexical items or grammatical rules). Strategy use varies considerably as a result of both the kind of task the learner is engaged in and individual learner preferences. (Ellis, 2008, p. 705)

Despite such guidelines (Ellis, 2008; Oxford, 1994) in defining strategy, it seems that the problems with the definition of strategy will persist in LLS research. LLS researchers will still use the definition(s) that are best for their research agenda. It is not the purpose of the present study to work out a definition of strategy that will embrace all LLS researchers’ agreement. The present study will adopt O’Malley and Chamot’s (1990) definition of LLS because they have consistently situated their research within cognitive theory, according to Anderson’s (1983, 1985, 1990, 2000, and 2005) Adaptive Control of Thought and Stages of Skill Acquisition Model.

Literature Review of Research on Language Learning Strategies

25

O’Malley and Chamot’s definition defined LLS as “special thoughts or behaviors that individuals use to help them comprehend, learn, or retain new information” (1990, p. 1). A detailed review of Anderson’s cognitive psychology theory in relation to LLS can be found in chapter 3. The review of LLS classifications above suggests that Oxford’s (1990) and O’Malley and Chamot’s (1990) classifications have been widely employed. However, Oxford’s classification is less suitable for the present study. The six categories of strategies can be effectively combined into three categories. For example, the memory strategies in Oxford’s (1990) are actually cognitive strategies. Therefore, it will be more parsimonious if the two categories (i.e., memory and cognitive strategies) in Oxford’s (1990) classification are subsumed under the category of cognitive strategies, as in O’Malley and Chamot’s (1990) cognitive strategies. In fact, Dörnyei (2005) and Dörnyei and Skehan (2003) have also proposed this change. Another issue with Oxford’s (1990) is the category of communication strategies. As mentioned earlier, the present study is not concerned with communication strategies. Communication strategies may or may not impact on learning (Cohen, 1998). In addition, O’Malley and Chamot’s (1990) classification is more theoretically sound given that it was developed based on Anderson’s (1983) Adaptive Control of Thought. This classification seems best fit for the present study. Therefore, O’Malley and Chamot’s (1990) tripartite taxonomy (see chapter 4) will be adopted in classifying the strategies used by Singapore primary school pupils. In summary, this section has revealed that differing LLS definitions and classifications exist in LLS research. Despite the differences, considerable progress has been made with respect to both definitions and classifications. Earlier researchers categorized LLS on the basis of their own observations and experiences. More recently, strategy classifications have been more data-driven (Chamot, 2004). Ellis (2008) also notes that there are now comprehensive, multileveled, and theoretically-motivated taxonomies (e.g., O’Malley & Chamot, 1990; Oxford, 1990) compared to the earlier ones (e.g., Naiman et al., 1978; Rubin, 1981; Stern, 1975) in which researchers just listed the strategies identified. LLS researchers have based their research on the definitions and classifications that they believe are most suitable for the types of LLS research they would like to pursue empirically, among which O’Malley and Chamot’s (1990) and Oxford’s (1990) definitions and classifications were the most widely adopted. For the present study, O’Malley and Chamot’s (1990) definition and classification of LLS will be adopted.

26

Chapter Two

Despite the ongoing debate on the definitions and classifications, two strands of LLS research have emerged. McDonough (1995) proposes a distinction between descriptive studies of strategy use and promotion and intervention studies, aimed at improving learners’ strategy use through strategy instruction. In the following section, I will mainly look at descriptive LLS studies. In section 2.3, LLS intervention studies will be reviewed.

2.2 Descriptive LLS studies Descriptive LLS studies in the literature have mainly examined the correlations between strategy use and students’ language learning achievements, with the latter being measured in various ways and covering different language skills. In my review of descriptive LLS studies, I will focus on those that investigated general LLS, LLS in specific skill areas, such as reading and listening, and their relationships to general language proficiency. As a great deal of LLS research concerning reading and listening has been conducted, patterns of strategy use have been established in these skill areas. The review of these areas will inform and benefit my study. As the focus of this study is to investigate writing strategies, I will then turn to research that focused on writing strategies. Currently, there is not much LLS research conducted in the area of speaking. In addition, this study views LLS as part of literacy strategies, i.e., reading & writing. Therefore, no review of speaking strategies will be touched upon.

2.2.1 The “good language learner” studies As most representative “good language learner” studies (e.g., Naiman et al., 1978, Rubin, 1975; Stern, 1975) were reviewed in 2.1.2, the same studies will not be reviewed in this section. LLS research began in the 1970s with “the Robin Hoodian good will of breaking the secret behavioral codes of successful language learners and sharing them with the unsuccessful ones” (Gu, 1996, p. 1). The “good language learner” studies by Naiman et al. (1978), Rubin (1975), and Stern (1975) set the cornerstone for the field of language learning strategies. These “good language learner” studies all aimed to identify strategies used by successful learners. Summarizing the research of “good language learners” by various researchers, such as Rubin (1975, 1981), and Naiman et al. (1978), Ellis (2008, p. 708) put forward five major aspects of successful language learning: 1) a concern for language form; 2) a concern for

Literature Review of Research on Language Learning Strategies

27

communication (functional practice); 3) an active task approach; 4) an awareness of the learning process; and 5) a capacity to use strategies flexibly in accordance with task requirements. Besides these five characteristics, there may be differences among the good language learners across contexts, cultures and countries. As pointed out in a “good language learners” investigation by Takeuchi (2003) as well as other researchers (e.g., Gu, 1996; LoCastro, 1994), significant differences exist among the so-called good language learners across contexts and cultures. At the same time, good language learners at different stages of their learning may have a varied repertoire of strategies (e.g., Griffiths, 2003, 2008; Takeuchi, 2003). Takeuchi (2003) found that the learners only started to use the strategy of “pushing oneself into using a foreign language” at their intermediate stage of learning. In Griffiths’s (2008) study, the good language learners were found to have the following strategies: 1) to manage their own learning; 2) to expand their vocabulary; 3) to improve their knowledge of grammar; 4) to involve the use of resources (such as TV or movies); and 5) to involve all language skills (reading, writing, listening and speaking). It is worth highlighting that the “good language learner” studies have generated rich insights into what successful language learners are like or do and how research can be carried out to investigate strategies that affect language learning (see Griffiths, 2015). However, these studies were mainly exploratory. One of the drawbacks is that no generalization can be drawn, mainly due to the small sample sizes and the research methods used. For example, these studies mainly used observations and interviews for data collection. Such methods can be used only when the sample size is small. Another drawback is that data collected through verbal reports might not represent the true picture of learners’ LLS use as some learners may be more willing to describe their learning strategies than others. Furthermore, some learners may be better at talking about language learning, whereas others are not. These issues were not addressed in the research on the “good language learner” and have underscored the need for more refined research methods and instruments with which to investigate LLS use in relation to language learning outcomes. Nevertheless, the aforementioned “good language learner” studies have inspired a considerable amount of LLS research (e.g., Chamot & ElDinary, 1999; Gu et al., 2005; Lan & Oxford, 2003; Magogwe & Oliver, 2007; Nong-Nam & Leavell, 2006; O’Malley & Chamot, 1990; Vandergrift

28

Chapter Two

2003a), comparing the patterns that effective and less effective language learners exhibited in their strategy use.

2.2.2 LLS and language learning outcomes The majority of the LLS studies have been devoted to investigating language learning strategies in relation to individual language learning achievements. These studies are generally descriptive. Besides the “good language learner” studies reviewed in the previous section, other LLS studies have explored the range and type of language learning strategies employed by effective language learners, the differences in LLS use between more and less effective language learners, and the relationship between strategy use and learners’ language learning outcomes. In this section, I will mainly review LLS studies related to language learning outcomes to identify patterns and useful findings to inform the present study. Chamot and El-Dinary’s (1999) study was part of a sophisticated longitudinal project investigating learning strategy application in elementary French, Japanese, and Spanish immersion programs. The participants were all native speakers of English. In the French and Spanish immersion programs, all the content areas were taught in the target language, whereas the students in the Japanese program only received Japanese instruction for certain subjects (e.g., mathematics, science and health). Data on two types of tasks (reading and writing) were collected through classroom observations, annual think-aloud protocols, questionnaires, and teacher interviews. No significant differences in overall measures of strategy use between the proficient and less proficient students were found. However, the thinkaloud protocols show that the proficient learners in this study reported using a variety of strategies for a particular task, whereas the less proficient ones were inflexible in orchestrating their strategy use, similar to Anderson’s (1991) findings on readers’ orchestrating of strategies. For example, on a reading task, the proficient learners focused more on sophisticated reading strategies, such as using background knowledge and making inferences. However, the less proficient ones relied heavily on phonetic decoding, which constituted 44% of their strategy use. An interesting finding is that all the students (effective vs. ineffective) employed about double the number of strategies for reading, compared with that for writing. In terms of strategy categories, for reading, more

Literature Review of Research on Language Learning Strategies

29

cognitive strategies were used by the students than the other categories, whereas for writing, metacognitive strategies were employed in greater proportion. The findings of this study confirmed those of other LLS studies (e.g., O’Malley & Chamot, 1990; Padron & Waxman, 1988; Vandergrift, 1997). Another investigation involving 168 third year Chinese students by Nisbet, Tindall, and Arroyo (2005) also examined the relationship between the use of LLS and English proficiency. Oxford’s Strategy Inventory for Language Learning (SILL) was used for data collection. The participants’ English proficiency was measured through the administration of Test of English as a Foreign Language (TOEFL). It was found that metacognitive strategies were the most frequently used strategies followed by social and cognitive strategies. The least used strategies were memory strategies. Among these strategies, only metacognitive were significantly but not highly correlated with their TOEFL results. Although a combination of metacognitive strategies and affective strategies were also found significantly correlated with the TOEFL results, they only accounted for 4% in variation. The findings suggest there was a weak correlation between LLS and the participants’ English proficiency. Similar findings were also reported in Chamot and El-Dinary’s (1999) study. The researchers (Nisbet et al., 2005) attributed the low correlation to the possibility that SILL did not include all the strategies used by the students. Both Chamot and El-Dinary’s (1999) and Nisbet et al.’s (2005) studies have brought up an issue with questionnaire-based research. It seems that, though useful, questionnaires alone may not be a very effective tool in eliciting LLS from language learners. In addition, adoption of an existing questionnaire such as SILL might not be feasible for the present study, given that the majority of the available questionnaires (e.g., SILL, Oxford, 1990) are meant for older adolescents and adults. Lan and Oxford’s (2003) large-scale investigation aimed to profile language learning strategies of elementary students in Taiwan. A strategy questionnaire was administered to 379 primary 6 students in a public primary school. The students were at three English proficiency levels. The questionnaire, adapted from Gunning’s (1997, cited in Lan and Oxford, 2003) Children’s SILL (based on Oxford’s 1990 SILL) was translated into Chinese with special considerations on simplification and comprehensibility in order to minimize any problem that young children may have. The analysis showed that the students’ language proficiency was significantly correlated with their overall strategy use. The high proficiency students (M

30

Chapter Two

= 3.2) used more strategies than the medium proficiency ones (M = 2.7), who in turn outperformed their low proficiency counterparts (M = 2.4). Language proficiency levels were also found to be significantly related to the use of cognitive strategies. A similar pattern to their overall strategy use was found. The high proficiency students used the cognitive strategies more often than their less proficient counterparts. The low proficiency group used the cognitive strategies least frequently. In Zhang, Gu, and Hu’s (2008) study, however, differences were found in the use of metacognitive strategies between the high proficiency and the low proficiency learners. The high proficiency learners used metacognitive strategies more frequently than their low frequency counterparts. Gu, Hu, and Zhang (2005) studied the think-aloud protocols collected from 18 pupils in a Singapore primary school. They found the high proficiency group employed more strategies and strategy use was qualitatively different. For example, one high proficiency pupil from primary 1 applied a wider range of strategies, including relating pictures to the story, anticipating the development of the story based on world knowledge, self-initiation, evaluation of understanding, predicting, and monitoring. However, a weak pupil from the same class only attempted to use such strategies as word-for-word repetition, ignoring problems, and abandoning effort to tackle a problem. The findings of this study confirmed those of Wharton’s (2000) study with 678 university students learning Japanese and French as foreign languages in Singapore. Data collected using SILL showed more strategy use among learners with higher proficiency (Griffiths, 2003; Lan & Oxford, 2003). In contrast, Loh (2007) found that primary 6 high proficiency pupils did not necessarily employ more learning strategies in another Singapore school. Whilst many LLS studies seemed to have discovered a linear relationship between frequency of strategy use and language learning outcomes, there are also different findings in the literature. Hong-Nam and Leavell (2006) undertook a study, examining the relationship between LLS use and second language proficiency. Their subjects were 55 ESL students studying in an American university. Two instruments, SILL (Oxford, 1990), and a background questionnaire designed by the researchers, were used for data collection. Overall, the students reported using more metacognitive and social strategies, as compared to other strategies, whereas the least used strategies were affective strategies and memory strategies. Interestingly, a curvilinear relationship was found between students’ strategy use and their language proficiency. Intermediate

Literature Review of Research on Language Learning Strategies

31

students reported using more strategies overall than their beginning and advanced peers. Similar results were also reported in another study (Philips, 1991). Philips, in her study with 141 adult ESL learners, discovered that there was no consistent difference in overall strategy use between the low proficiency and the high proficiency learners. However, the intermediate proficiency learners showed more strategy use on SILL. Ikeda and Takeuchi (2003) looked at the differences of strategy use between five proficient readers and five less proficient readers in Japan. The researchers collected portfolios made by the 10 students, as well as their reflections on their strategy use. Significant differences were found between the high proficiency readers and their low proficiency counterparts. The proficient readers not only reported more use of reading strategies but also were good at orchestrating their strategies, as compared to the less proficient readers, in agreement with the findings of other studies (e.g., Anderson, 1991; Chamot & El-Dinary, 1999; Chamot & Küpper, 1989; Hong-Nam & Leavell, 2006; Philips, 1991). In comparison with Ikeda and Takeuchi (2003), Oxford, Cho, Leung, and Kim (2004) explored the effects of task difficulty and strategy use, involving 36 ESL university students who were general academic majors in the US. For the easy reading task, there was little difference in strategy use between the proficient and less proficient readers, whereas for the more difficult task the less proficient readers employed more reading strategies than the proficient readers. The interpretation provided by the researchers is that the proficient readers did not have to use many learning strategies to do the task because the “difficult task” was not much of a challenge for them. The sample size of most of the aforementioned LLS studies was quite small. In contrast, Magogwe and Oliver (2007) undertook an investigation with 480 students from three levels of schools, i.e., primary schools, secondary schools and a tertiary institution. A modified version of SILL was used for data collection on strategies. The analysis showed that there existed differences associated with the proficiency levels of the students within and among the three levels. For example, primary school students reported more social strategy use than the other categories. However, the most frequently used strategies for both the secondary school students and the tertiary students were metacognitive strategies. Overall, the high proficiency students used more strategies than their counterparts. At the primary school level, the high proficiency students preferred social and

32

Chapter Two

metacognitive strategies as compared to the low proficiency students. The high proficiency secondary school students used more metacognitive, social, cognitive, and compensation strategies than the low proficiency ones did. At the tertiary level, the high proficiency students reported more use of metacognitive strategies than the low proficiency students. The findings of this study are also consistent with those of other studies (e.g., Gu et al., 2005; Ikeda & Takeuchi, 2003; Oxford & Ehrman, 1995; Yang, 2007; Zhang & Xiao, 2006) in that high proficiency students employ LLS more often than their low proficiency counterparts. The earlier part of this section mainly focused on the descriptive studies investigating differences between effective and less effective language learners in their overall strategy use as well as the relationships of LLS to proficiency in the target language. In the following part, I will review mainly descriptive studies with respect to specific LLS. Similar to the research reviewed in the previous section, research concerning specific language learning strategies investigated the range and type of LLS in relation to a particular language skill such as reading and listening, and differences between effective and less effective language learners in their strategy use. As it is impossible to cover all the studies in relation to specific LLS, e.g., reading, listening and writing, only a few representative ones will be reviewed here. As reviewed above, Gu et al. (2005) not only found differences between good and weak pupils in their overall strategy use, but also strategy use for reading and writing. In their study, the majority of the high proficiency pupils used more effective strategies such as predicting the development of the reading text, verifying their anticipation, and reinterpreting the information when necessary. In writing, they frequently used planning, rereading, and monitoring (grammar as well as spelling) strategies. In contrast, the weak pupils focused more on bottom-up processing strategies on the reading tasks. Seldom did they plan before writing. They only monitored their spelling mistakes when writing. Zhang, Gu, and Hu (2008) focused on the reading strategies used by primary school pupils. The overall findings of this study corroborated those of Gu et al. (2005). Significant differences were found between the high proficiency and the low proficiency pupils not only in their overall strategy use but also in their use of individual strategies. In terms of metacognitive strategies, the high proficiency pupils showed a higher frequency of use than their low proficiency counterparts. Specifically, the

Literature Review of Research on Language Learning Strategies

33

high proficiency pupils used more self-initiation, planning, and monitoring. Also, the two groups differed significantly in their use of three cognitive strategies: decoding, inferencing, and prediction. The low proficiency pupils preferred to use decoding more frequently, whereas the high proficiency pupils used more inferencing and prediction. As discussed by the researchers, the difference between the two groups was not only quantitative but qualitative. Although both groups of pupils used decoding, the low proficiency group was found to overly rely on this strategy as compared with the high proficiency group. The high proficiency group employed this strategy only when they saw the need. Moreover, the high proficiency group did not restrict their comprehension to the local details, e.g., words and/or phrases. Despite the differing results produced in the area of reading strategies, research shows that students do use a variety of reading strategies that facilitate their reading comprehension. Anderson (1991) notes that strategic reading is not only a matter of knowing what strategies to use, but the reader must know how to use a strategy successfully and orchestrate its use with other strategies. This has also been echoed by many LLS researchers (e.g., Zhang et al., 2008). It is not sufficient to know about strategies; the reader must also be able to apply them strategically (e.g., Ikeda & Takeuchi, 2003, 2006). O’Malley, Chamot, and Küpper (1989) examined five effective and three less effective high school students’ L2 listening strategies. They found that the effective listeners were able to monitor and direct their attention. However, the less effective listeners were easily carried away from the listening task by unknown words or phrases, unaware of their inattention. The effective listeners employed top-down processing strategies mostly, as compared to their less effective peers who used bottom-up processing strategies exclusively on the task. Vandergrift’s (2003a) study reported similar findings about 36 junior high school students in Canada. The results from this study indicated that the skilled listeners were good at controlling the listening process, planning for the task, and monitoring incoming input, whereas the less skilled listeners lacked comprehension monitoring and used on-line translation and bottom-up processing strategies. The research findings suggested that the skilled listeners tended to use their listening strategies in an orchestrated manner, as reported by Ikeda and Takeuchi (2003). Besides the listening strategies reported in the above literature, Goh’s (2002) study identified two new strategies: fixation and real-time assessment of input, which were rarely found in previous

34

Chapter Two

research. Fixation was shown to be counterproductive, whereas real-time assessment was found to be facilitative in listening comprehension. Gu, Hu, and Zhang (2004) confirmed most of the findings in the previous research on listening strategies (e.g., Goh, 2002; O’Malley et al., 1989; Vandergrift, 2003a). The study involved 18 pupils in Singapore primary schools. The analysis of the think-aloud protocols suggested that the good listeners exhibited more regulation in metacognition and better metacognitive awareness than the poor listeners. Cognitive strategies used by the good listeners tended to be meaning-oriented listening strategies (e.g., inferencing, predicting, and elaborating), whereas the poor listeners relied heavily on perceptual processing and bottom-up listening strategies. In sum, the literature reviewed above has shown a complex relationship between strategy use and language proficiency. Although many studies showed that students with high proficiency used more LLS, some others found different relationships. Nonetheless, the majority of LLS researchers agree that appropriate use of language learning strategies helps enhance learners’ language learning. Both effective and less effective learners use LLS to a certain extent, but they use them differently. Effective language learners are skilled at matching language learning strategies to the tasks and flexible in applying learning strategies. In other words, effective orchestration of strategies characterizes successful language learning (Anderson, 2005; Macaro, 2009; Zhang et al., 2008). However, less effective learners use LLS in a more rigid manner and tend to focus only on a certain group of strategies, e.g., decoding in reading all the time.

2.2.3 Research on writing strategies Research on L2 writing strategies originated in and benefited from L1 writing studies (e.g., Flower & Hayes, 1980, 1981). The term used in the literature to refer to studies of L1 writing strategies is “writing process research”. Therefore, the research of writing strategies has been part of research on process-oriented writing. However, L2 writing strategies have been under-researched, as compared to the abundant LLS research in relation to language learning outcomes as reviewed in 2.2.2. Writing strategies involve the writers’ employment of particular processes or techniques in order to enhance their writing. Empirical studies of writing strategies were inspired by the theory of recursive writing processes by Perl (1979) and others (e.g., Flowers & Hayes, 1980, 1981)

Literature Review of Research on Language Learning Strategies

35

in contrast to linear stage theory of writing. The theory of recursive writing postulates that writing is comprised of a series of recursive processes, whereas linear stage theory contends writing follows a strict plan-write-revise sequence. The problem with stage models of writing is that they model the growth of the written product without looking at the inner processes that the writer is engaged in. Flower and Hayes (1981) put forward their writing model which involves three primary cognitive processes: planning, translating, and reviewing. This line of research in L1 studies has contributed valuable insights into writing processes by showing that expert writers differ significantly from novice writers in their composing processes. To be specific, Perl found that both the skilled and unskilled writers discovered their ideas in the process of composing. However, the unskilled writers were easily carried off-track during the process of exploring these ideas due to their surface-level concerns. Moreover, many process-based writing researchers found that competence in the composing process was more important than writers’ linguistic competence in order to write well (e.g., Jacobs, 1982; Jones, 1982; Zamel, 1982, cited in Krapels, 1990). ESL teachers and researchers, influenced by the L1 composing processes research, also shifted their attention from product to process-centered teaching of ESL compositions. Zamel (1983) examined the composing processes of six advanced ESL students. This study adopted a case study approach, using student interviews and observations as major data collection methods. Although the ESL students were observed while composing so that both their writing behaviors and what they wrote could be recorded, they were not required to verbalize their thoughts (e.g., to think aloud). Zamel was concerned that the think-aloud procedure might change the students’ thinking processes. Each student was given a courserelated writing task and allowed to complete it at his/her own pace. The findings, similar to those in first language writing research, indicated that skilled writers perceived writing as a non-linear process of discovering and exploring ideas and constructing a framework to present their ideas. Influenced by writing process research, several LLS studies looked at specific writing strategies employed by L2 writers (e.g., Perl, 1980; Raimes, 1985; Victori, 1999; Zamel, 1983). Like research on strategies in other skill areas, these studies also aimed to identify the differences between more and less effective writers in their use of specific writing strategies as well as orchestration of strategy use. Their findings indicated that less effective writers, though aware of the availability of writing

36

Chapter Two

strategies, only focused on the written text per se. Very little or insufficient time was spent in planning and revising. Less effective writers perceived writing as somewhat linear processes. However, their more successful peers did substantial planning before writing and were more concerned with global structures than surface issues (e.g., grammar and vocabulary). They also attempted to deploy more writing strategies while writing. To them, writing was comprised of recursive and non-linear processes. Besides finding out skilled writers’ perceptions on writing processes, Zamel’s (1983) study also looked at the planning strategies of the six advanced ESL students, who were completing a writing task. Zamel found that the four skilled writers viewed their planning as a non-unitary stage. They were aware that they could leave their thoughts and revisit them later. However, the two unskilled writers viewed writing as immediate and straightforward expansion of their plan. For example, when the least skilled writer in this study was stuck during planning, she asked the researcher, “Can I add something later?” Clearly, the unskilled writer was not aware that “planning is not a unitary stage, but a distinctive thinking process which writers use over and over again during composing,” (Flower & Hayes, 1981, p. 375). Raimes (1985) investigated eight unskilled ESL university students’ use of specific writing strategies by analyzing their think-aloud protocols. The students were asked to think aloud their writing processes while writing a narrative. Raimes found that some of the students did not spend much time in their pre-writing activities (e.g., reading the topic, rehearsing, planning, trying out beginnings, and making notes). They were more focused on producing the text as compared to their other writing strategies. This finding is consistent with Zamel’s (1983) finding on the least skilled writer. As regards revising and editing, they were relatively more concerned with idea generating than with surface changes. She argued that ESL students in writing classes should not be treated simply as learners who need large doses of language input to improve their writing. Less skilled writers need “more instruction and practice of generating, organizing, and revising ideas; more attention to the rhetorical options available to them” (Raimes, 1985, p. 250). Compared with Zamel (1983) and Raimes (1985), Victori (1999) investigated a greater range of writing strategies, e.g., planning, organizing, and evaluating (revising and editing) that two effective and two less

Literature Review of Research on Language Learning Strategies

37

effective writers used. The data consisted of the writers’ think-aloud protocols and interviews. The researcher found that the effective writers were engaged in some kind of planning before writing and were not constrained by their initial plan when writing. With respect to planning strategies, this study confirmed the findings of Gu et al.’s (2005) study. However, the two less effective writers treated the writing process as speaking. They preferred to improvise and write as if they were speaking, similar to the ineffective writers in Raimes’ (1985) study. The two effective writers were concerned with organizing their ideas, in contrast to the two less effective ones. As regards evaluating, the two less effective writers attended to only surface level features of writing (i.e., grammar and vocabulary), but the two effective ones showed a broader range of evaluating strategies and were not limited to the surface-level features of writing. They tried out two other strategies, i.e., retrieval of new ideas and evaluation of linguistic structure. It is interesting to note that the two less effective writers in Zamel’s (1983) study perceived planning similarly to the two in this study. These less effective writers perceived planning as a process similar to speaking. The study by Chamot and El-Dinary (1999), reviewed earlier, also examined use of writing strategies by native English-speaking children studying in immersion programs in their third or fourth year of primary school. Planning was found to be the most frequently used strategy type by both effective and less effective writers. When writing, the less effective students seemed to focus more on details, whereas the effective ones were more concerned with the task as a whole. Similarly, Gu et al.’s (2005) study also showed ineffective learners’ exclusive focus on local details. Baker and Boonkit (2004) examined the writing strategies used by 149 Thai university students (successful and less successful writers). The data were collected through a questionnaire, learning journals, and interviews for data triangulation purposes. The most frequently employed categories of strategies were metacognitive, cognitive, and compensation strategies. The most frequently used metacognitive strategy (M = 4.36/5) was “I use my background (world) knowledge to help me with ideas”. The most often used cognitive strategy was “I often write in Thai”. The least used metacognitive strategy (M = 1.89/5) was “I make a timetable for when I will do my writing”. However, the results did not show any significant differences between the successful and less successful writers in their use of the six categories of strategies on Oxford’s (1990) SILL. Only the “negative” strategies on the questionnaire, including strategies that hinder

38

Chapter Two

writing process, were found to be used differently by the successful and less successful writers. Macaro (2007) studied writing strategies employed by 16 beginner learners of French. The data were collected through task-based interviews with the participants. The analysis found that all the participants relied on “copy-writing” as their writing strategies. They copied from their textbooks, from the whiteboard or from worksheets. The strategies identified were mainly cognitive. For example, the students tried to make sense of meaning while copying. Some students reported that they looked for meaning first in their copying. Encoding the language from written form into sounds was also widely used by these students. They also used “avoiding” and “monitoring” strategies. For example, when they could not produce a desirable phrase, they would avoid saying the idea. In comparison with Baker and Boonkit’s (2004) study, Macaro (2007) found different writing strategies with his participants. One possible reason may be that his participants were only beginner L2 learners, while Baker and Boonkit’s (2004) participants had studied the L2 for a few years. Another reason may lie in the fact that Macaro (2007) used task-based interviews for data elicitation. Such a data collection method tends to elicit more online writing strategies. Chien’s (2012) study involved 40 year-2 students, who majored in foreign languages and literature in a university in Taiwan. Through think-aloud protocols and retrospective interviews, significant differences between the 20 effective and 20 ineffective writers were found. The effective writers used their global planning, text-generating, revising and editing text more often than the 20 ineffective writers. However, the effective writers did less planning when they interpreted the writing task. Chien’s explanation was that the effective writers may have known a lot about what to write, thus spending less effort interpreting the writing task. The studies of writing strategies reviewed above mainly involved ESL students doing writing tasks in their language courses. These studies provide insights into how these students employed and benefited from some of the writing strategies in their writing classes. It is also beneficial to examine studies on ESL students’ use of L2 writing strategies in disciplinary courses. Leki (1995) investigated coping strategies of five ESL university students for their writing tasks in disciplinary courses in an American university. The participants of this qualitative study were three graduate and two undergraduate students who had done quite well in their

Literature Review of Research on Language Learning Strategies

39

disciplinary courses and their writing tasks in the previous semester. The participants differed in sex, age, home country, year in school, and majors. The data were collected from four main sources to ensure triangulation of information: interviews with the students and their professors, observations, and the students’ artifacts. Ten broad strategy categories were identified: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

classifying strategies focusing strategies relying on past writing experiences taking advantage of first language/culture using current experience or feedback looking for models using current or past ESL writing training accommodating teachers’ demands resisting teachers’ demands managing competing demands

This study sheds some light on ESL students’ use of writing strategies in disciplinary-specific courses, showing that some students had been equipped with some well-developed strategies prior to their studies in a university. They were able to transfer some of the strategies learned previously to the writing tasks across the curriculum. Their previous learning experiences (in their L1 context) in their home countries provided them with strategies to cope with their studies in their L2. Besides, these students could deploy their strategies flexibly to achieve their desired outcomes. For example, when Ling, a participant in the study, failed to do well in her studies through employing her often-used strategy of “going to the library and reading books for information” that had been developed in her past learning experience, she was able to alter her old strategy and try new ones, including appealing for help and appealing for clarification. Leki (1995) noted it would be a good practice to train ESL students in the English for Academic Purposes (EAP) courses to use some of the LLS that successful ESL students already use for their future encounters with writing assignments across the curriculum. A few strengths with regard to the above-reviewed studies are worth noting. Investigating specific writing strategies that L2 learners use can lead to insights into how individual learners (effective vs. ineffective writers) differ in their choice and use of writing strategies. For example, it shows that planning is a strategy associated with effective writers (e.g.,

40

Chapter Two

Chamot & El-Dinary, 1999; Chien, 2012; Gu et al., 2005; Victori, 1999; Zamel, 1983). Ineffective writers either do not plan or view planning as a linear process. Effective writers perceive writing as a recursive process, are more concerned with overall organization and idea generating, re-read and monitor their writing processes and use more evaluating strategies, while ineffective writers tend to move from planning to writing to revising without recursively working on their drafts, and focus more on local details, e.g., spelling, when writing, and only make surface changes to the text when revising. Also, this line of research can provide teachers and researchers with a basic understanding of learners’ writing processes for their effective classroom instruction. One of the major advantages of identifying the types of strategies writers use is that researchers can compare the writing strategies that good and poor writers use under the theory of process writing. Unlike a stage model of writing, the major analysis units of which are the stages of completion, reflecting the growth of a written product in a linear sequence, a process model offers more aspects, through which the act of writers’ writing can be examined. Of course, it would be ideal to try to firstly identify some universal writing strategies that can be applied to all writers, regardless of their differences. Besides, many studies in this line of research employed case studies and writers’ think-aloud protocols analysis, which represents efforts into looking at what writers were actually doing when they were writing (Grabe & Kaplan, 1996). This has contributed greatly to the field in general.

2.2.4. Critique of descriptive research on LLS In section 2.2.2, I reviewed the relationship between the use of LLS and language learning achievements. The LLS research in this area aimed to identify the relationship between strategy use and learners’ language proficiency as one of the research goals. However, it seems that not all the findings showed the same patterns. As concluded by McDongouh (1999) in his then state-of-the-art review of LLS research, “the relationship between strategy use and proficiency is very complicated: Issues such as frequency and quality of strategy use do not bear a simple linear relationship to achievement in a second language” (p. 13). For example, although a large number of studies show that proficient learners employ more language learning strategies (e.g., Lan & Oxford, 2003; Magogwe & Oliver’s, 2007; Zhang et al., 2008), other studies (e.g., Hong-Nam and Leavell, 2006; Philips, 1991) indicate that it is the students with moderate

Literature Review of Research on Language Learning Strategies

41

proficiency that use more strategies than students at the other proficiency levels. In addition, no significant correlations were found in Nisbet et al.’s (2005) study. Second, there is also a complex relationship between students’ language proficiency levels and types of strategies used. The most often used are cognitive strategies among Sugeng’s (1997) Indonesian primary school students. However, Lan and Oxford’s (2003) findings suggested that the most frequently used strategies among the 379 Taiwanese students are compensation and affective strategies. Chamot and El-Dinary’s (1999) study presented an even more complicated picture. For reading, more cognitive strategies were frequently used by the students than the other categories. Metacognitive strategies, however, were most often employed by the students while writing. It is not easy to explain the differences found in these studies. The varied findings may have resulted from different research methods used in these studies. For example, Chamot and El-Dinary’s (1999) study analyzed the participants’ think-aloud protocols which revealed that the proficient students were better at orchestrating their strategies than the less proficient ones. However, their questionnaire survey found no significant differences between the two groups. Third, some studies (e.g., Raimes, 1985) relied on think-aloud protocols as their main data collection method. The singular research method can only reveal certain aspects of writing strategies of the learners as a result of the particular method. Only task-specific writing strategies may be revealed through the think-aloud protocols. At the same time, the think-aloud protocols tend to reveal only online processing of the learners, rather than metacognitive aspects of planning or monitoring (Chamot, 2004). Also, given the small sample size used in case studies, it is difficult to generalize the research findings to a larger population. As such, other data collection methods, such as questionnaires, learning journals, interviews, students’ written texts, and simulated recall protocols should also be used. For example, questionnaires may elicit a wider range of strategies (e.g., Chamot & El-Dinary, 1999; Lan & Oxford, 2003; Magogwe & Oliver, 2007; Nisbet et al., 2005). The fourth problem with the above-reviewed studies is the depth of analysis that was reached. The majority of the studies employed closedresponse questionnaires such as SILL as data collection methods and relied on quantitative data analysis (e.g., Lan & Oxford, 2003; Magogwe & Oliver, 2007; Nisbet et al., 2005). The benefits of this type of research

42

Chapter Two

design are quite obvious: large sample size, relatively easy data collection, and quantitative analysis by computer software. However, some other studies which did not (only) rely on questionnaire surveys revealed a recurring finding that less effective learners do employ language learning strategies, sometimes as frequently as their more successful counterparts. It may not be the quantity that matters, as discussed by Tseng, Dörnyei, and Schmitt (2006). For example, Chamot (2005) noted that less proficient learners’ deployment of LLS is ineffective as a result of lack of metacognitive knowledge with regard to task requirements. On the other hand, effective language learners are good at selecting learning strategies appropriate to the task they are engaged in. However, the complex interactions between strategies and tasks may not be fully captured through questionnaires. For example, Ehrman, Leaver, and Oxford (2003) discovered that less proficient learners often used strategies in a random manner. Questionnaires have largely failed to address such qualitative aspects. In future research, a combined research method (e.g., questionnaires, think-alouds and interviews) should be employed for effective data collection. Although there have been concerns with employing think-alouds with primary school students, research has shown young primary students were able to describe their own thinking and learning processes (e.g., Chamot & El-Dinary, 1999; Gu et al., 2005). There were also similar findings from other LLS studies (e.g., O’Malley & Chamot, 1990; Padron & Waxman, 1988; Vandergrift, 1997). Fifth, as L2 writers come from diverse backgrounds in terms of their cultures, L1s, and previous writing experience, an exclusive focus on supposedly uniform cognitive processes of the writers may not sufficiently account for what makes the difference between proficient and less proficient writers. Intuitively, not all writers undergo identical composing processes and possess same cognitive capabilities (Grabe & Kaplan, 1996). Atkinson (2003a, 2003b) points out that writing is not a decontextualized, information-centered, impersonal activity, but it is rather informed by and situated within power relations, society, and individuality itself. Thus, perceiving writing only as cognitive processes or exclusively from a process-writing perspective is insufficient and will not lead to more effective writing instruction. The socio-cultural dimension of writing plays an important role in helping improve students’ literacy. A socio-cultural view of writing is not to discredit process writing but rather to perceive writing processes as socio-culturally situated. In this regard, many researchers have looked at, among other research agendas, how peer review/teacher feedback can help learners improve their writing from a

Literature Review of Research on Language Learning Strategies

43

socio-cultural perspective (e.g., Bitchener, Young, & Cameron, 2005; Ferris & Roberts, 2001; Hu, 2005; Min, 2006, Urzua, 1987, for a detailed review, see Fitzgerald, 1987). From a cognitive perspective, L2 writers go through a number of procedures: pre-writing, planning, writing and rehearsing. These procedures can be said to constitute the cognitive dimensions of literacy learning in the light of the multiple-dimensional view of literacy (Kucer, 2005, 2009). In this view, literacy learning constitutes such multiple dimensions as cognitive, linguistic, socio-cultural and developmental dimensions. All these dimensions are essential ones for students to learn. Therefore, it can be argued that learning also takes place in the other dimensions of literacy, i.e., the linguistic dimension, the socio-cultural dimension, and the developmental dimension in writing. A review of Kucer’s (2005, 2009) view can be found in chapter 3. To sum up, LLS researchers have arrived at different conclusions with regard to strategy use and proficiency. However, the research is still of significance as it offers insights into learners’ strategy use in relation to their language proficiency, which helps broaden our understanding of the characteristics of individual learners in relation to their LLS use. Therefore, it will be interesting to identify the strategies used by Singapore primary school pupils and investigate the relationship between their LLS use and proficiency.

2.3 Intervention research on LLS As reviewed in section 2.2, most LLS investigations have been simply descriptive, committed to identifying, describing and classifying learning strategies used by language learners. Once the strategies used by successful learners are known and the strategy use of successful and less successful learners is compared, the follow-up task is to teach those less successful learners how to use LLS effectively so that they can improve their use of LLS and language competence. The underlying premise of learner training is that successful teaching of those strategies used by effective learners to less effective learners can help enhance their language learning. It is this premise that has given rise to intervention studies of LLS. These studies have also generated mixed results. Several strategy instruction frameworks have been used in strategy-based instruction, which is a learner-centered approach to teaching strategies. Though the details vary, they have similar objectives (see Cohen, 1998; Harris, 2003).

44

Chapter Two

These objectives include 1) awareness-raising as to the purpose and rationale for strategy use; 2) practice of learning strategies being taught; and 3) strategy application in new learning contexts. In this section, I will review intervention studies, which can be further divided into studies (2.3.1) aimed at teaching general strategies for overcoming language learning problems, and studies (2.3.2) that are aimed to teach skill-specific strategies, e.g., reading strategies, listening strategies, and writing strategies.

2.3.1 Intervention studies of general LLS The general LLS intervention studies to be reviewed in this section were undertaken under diverse conditions with varying research agendas. Some studies built on and followed from earlier descriptive studies. Some embarked on strategy training directly, using various methods. However, these general studies are still of relevance to the present study as they can serve as either good examples to follow or unsuccessful ones to be reflected upon. Wenden (1986) conducted an intervention study with 23 students studying in a seven-week intensive English course to raise their awareness of strategy use. Out of the 20 class hours each week, two hours was used for strategy training. The training tasks consisted of comprehension exercises, class discussions based on the reading or listening passages, out-of-class practice tasks, and focused diary writing. The tasks were adapted from materials for language learning. At the end of the training, a strategy questionnaire was administered to evaluate the effectiveness of and students’ attitudes towards the training. Wenden found that the students involved were generally not interested in the training and saw it as irrelevant. She described the students as “resistant” to the training. Wenden’s explanation of students’ disinterest was that the strategy training should have been more integrated into a particular language task or objective. By revising the materials, the training was found to attract more students’ interest subsequently. However, she did not measure the effectiveness of the training after the modification of materials. This study was one of the several pioneer intervention studies in the literature. While Wenden’s study did not focus on particular language tasks, O’Malley et al. (1985b) undertook a two-phase study on LLS using specific tasks. Phase 1 of the study aimed to identify the range and frequency of learning

Literature Review of Research on Language Learning Strategies

45

strategies used by high school ESL students. Phase 2 aimed to determine the effects of LLS training on the students. The participants were 70 high school intermediate-level ESL students with various L1 backgrounds. The objective of the intervention was to test whether LLS instruction in a natural setting could help enhance ESL students’ learning for integrative second language tasks (e.g., listening, speaking, and vocabulary learning). The researchers compared the improvement on the language tasks made by three groups of learners: a group that received combined instruction on metacognitive, cognitive, and social/affective strategies, a group that received instruction on cognitive and social/affective strategies, and a control group that received no training and worked on their language learning tasks the way they normally did. Analysis of the study identified significant statistical differences among the three groups for the speaking task with the first group outperforming the second group, which in turn scored higher than the control group. For the listening task, the results were mixed. Significance was found for the human interest stories, the more interesting and less demanding of the tasks, whereas no significance was found for the geography or science lessons. On the vocabulary test, no significance was found. On the one hand, the researchers noted that the Asian students in the control group preferred to use rote repetition. On the other, the Asian students in the treatment groups resisted the training (O’Malley & Chamot, 1990). This might have contributed to lack of statistically significant differences between the groups. In contrast, the Hispanic students in the treatment groups performed better than those in the control group. One possible interpretation for this is that the cultural background mediated the students’ learning during the training. Politzer and McGroarty (1985) note that cultural background generally influences students’ learning behaviors and teachers find it a challenge to match their teaching styles to students’ learning behaviors. This pioneering study of LLS training laid the groundwork for subsequent research on strategy training. However, the effects of the training on metacognitive, cognitive, and social/affective strategies were quite mixed. Cohen (1998) commented that it is difficult to measure the effects of explicit strategy training if learners’ report of strategy use is not linked to performance on specific tasks. LLS are deployed depending on particular tasks that can vary in many ways (Chamot, 2005). Although some findings of the study indicated that classroom instruction in LLS with integrative language skills could help improve learning, we

46

Chapter Two

should be cautious with respect to the effects. First, it was the researchers rather than classroom teachers that provided explicit strategy instruction to students in this study. Further research needs to involve real classroom teachers to determine the effects of strategy training in a natural classroom setting (Chamot, 2005; Gu, 1996), as teachers play an important role in managing the classroom ecology (Hawkins, 2004). Second, the effect with the listening task was not strong enough for a clear-cut conclusion on the effectiveness of strategy training on the listening task. The researchers suggested that the weak effect was due to the difficulty of the learning task. A pilot study to test the difficulty level of the materials might help avoid this problem. Third, although the researchers did not specify it, it seems that the treatment students received daily 50-minute strategy instruction for only eight days. The short duration of the intervention and the rather concentrated training sessions might not have provided the students with enough time and practice to hone their use of strategies newly learned. In contrast to the above studies in terms of design, Nunan’s (1997) study was better controlled. The study adopted an experimental design investigating the LLS training effects with 60 first year students in a Hong Kong university. The participants were randomly assigned to four different classes: two experimental and two control classes. All the four classes took the same course (English for Arts). Only the two experimental classes were systematically taught learning strategies associated with the course during the semester. Before the training, a questionnaire assessing their knowledge, use, and perception of the value of 15 strategies was administered to all the four classes. Upon the completion of the course, all the students took the same questionnaire again. The results showed high increases in the experimental students’ knowledge of the strategies and their perception of the value. However, no difference was found in their overall use of the strategies taught between the experimental and the control classes. In fact, there was a dramatic increase in strategy use for all the classes. Nunan attributed this finding to the fact that the first year university students were expected to be much more independent and selfdirected as compared with their own high school life. Therefore, they had to develop a range of strategies to cope with their new way of learning and living. A strategy training study involving classroom teachers as instructors is reported in chapter 6 of O’Malley and Chamot (1990). This study was part of a larger project consisting of two other studies: a descriptive study (Chamot et al., 1987) reviewed earlier and a longitudinal study. The main

Literature Review of Research on Language Learning Strategies

47

objective was to discover whether and how foreign language teachers would incorporate strategy training in language classes, using the materials and approaches chosen by the teachers based on their own judgments. The foreign language teachers participated on a voluntary basis. The researchers discussed typical learning strategies of the students, which were identified in their descriptive study, with the teachers in their meetings before the nine observations they made of the classes on listening, reading comprehension, and oral skills. For each observation, each instructor briefed the researcher with the lesson objectives, student needs, and instruction concerns. Time permitting, a debriefing session was held to explain issues that the researcher would like the teachers to know after each observation. The observations showed that the foreign language teachers were able to incorporate strategy training in their foreign language classes. This study is of particular relevance to the present study as it contains a descriptive part that led to the intervention phase, which informs the design of the current study. At the same time, it was different from the intervention studies reviewed previously in several ways. First, this study was aimed at determining the teachers’ effectiveness in incorporating strategy training in their foreign language class. In this study, the teachers taught the strategies. Second, the teachers were at liberty to select strategies they would like to teach as part of their regular foreign language classroom activities, whereas in O’Malley et al.’s (1985b) study, a whole 50-minute instruction session was dedicated to teaching students to use the strategies identified in the descriptive part of the larger project, which were selected and instructed by the researchers. O’Malley and Chamot (1990) noted that if SBI is to become an integral part of second language teaching, classroom teachers need not only to see the value of SBI but develop necessary instructional skills. In this regard, Chamot (2005) has also called for inclusion of classroom teachers in SBI. She has noted that the relationship between impact of LLS training and teacher characteristics, such as attitude, belief and LLS training, should be investigated. Therefore, more research studies involving classroom teachers should be conducted to investigate the effectiveness of SBI. Third, this study documented the way that the instructors implemented SBI and the types of strategies taught to the students, which is particularly relevant and useful to this study. Fourth, as explained by the researchers, some teachers’ nonparticipation resulted from their lack of interest in LLS teaching and lack of time for planning and instruction. Motivation of the teachers was also an important factor leading to successful implementation of SBI. The more motivated

48

Chapter Two

instructors implemented the training successfully and embraced the training with apparent enthusiasm, whereas the less motivated teachers showed indifference to it and gave up the training eventually. In this study, no teachers elected writing lessons for undertaking their strategy instruction. It would be interesting to find out the reason. Was it the skill (writing) itself or other reasons that made them prefer reading, listening, and speaking to writing? Kuba (2002) carried out an SBI study with 188 students from four classes in three high schools in Japan. Two classes were randomly assigned to the experimental group, the other two the control group. She used Oxford’s (1989) SILL to assess the students’ frequency of strategy use. The SBI lasted 14 weeks in which three 10 minutes’ sessions each week were given to the experimental classes. The classroom teachers were trained to carry out the SBI. Both the teachers and the researcher designed the SBI lesson plans. The results showed that all the experimental classes improved on their reported strategy use after the SBI. However, their strategy use was not significantly correlated with English achievements across all the three schools after the intervention. The results are consistent with those of Gu’s (1996) study in that improvement of strategy use on the questionnaire may increase after SBI, whereas it may not be correlated with language learning achievements.

2.3.2 Intervention studies of skill-specific LLS The above reviewed intervention studies were mainly concerned with SBI of general LLS. They pioneered and deepened our understanding of strategy-based training. Different from the intervention studies of general LLS, intervention studies of skill-specific strategies examine the effect of training students to use strategies in particular skill areas, for example, reading and writing strategies. In this section, I will mainly review intervention studies of skill-specific LLS. Carrell, Pharis, and Liberto (1989) conducted a small-scale intervention study, in which two metacognitive reading strategies were compared with each other with regard to their effects on two groups of students. The participants of the study were 26 ESL students aged between 19 and 43. Although the sample was small, the study assigned two intact reading classes of nine students to two experimental conditions. One experimental group received training in the use of semantic mapping, while the other group was taught to use an experience-text relationship method (ETR).

Literature Review of Research on Language Learning Strategies

49

The only control group comprised eight students. All the participants were pre- and post-tested on the same three reading comprehension passages. Both of the experimental groups outperformed the control group in the post-test after the four-day training, but different measures of reading showed different improvements. Although no difference was found for the multiple choice questions, the two experimental groups scored higher than the control group for the open-ended questions. The results showed that metacognitive training was effective in helping students improve their second language reading. The effectiveness of strategy training varied depending on the strategies in question, the training methods (e.g., semantic mapping training and the experience-text relationship method), and the measures used. Carrel et al.’s study has shown that metacognitive reading strategy training can be effective. This study showed some strengths as compared with other SBI studies. First, prior to the formal training, a pilot study was carried out to test the research design, as well as the materials and procedures. Second, both a pre- and post-test were conducted to measure the differences between the experimental groups and the control group. However, we should interpret the results with caution, as suggested by the authors, due to the rather small number of participants involved in the training, the same test used for both the pre- and post-test, and the short training period as well. Carrier (2003) carried out an exploratory intervention study of training in listening strategies with seven high school ESL students in the US. This study included two pre-tests and two post-tests. There was no control group. Fifteen training sessions of 20 to 30 minutes were used by the researcher to provide listening strategy training. The strategies taught included both top-down and bottom-up processing strategies, for example, selective attention to stress patterns and guessing from the context. The results showed that the students’ performance in discrete listening, video listening, as well as note-taking significantly improved. However, the research design prevents the researcher from claiming if the strategy training was successful since there was no control group. Another study (Vandergrift, 2003b) looked at the effects of awareness raising of the process of listening through tasks designed to develop effective listening strategies on 41 university students, who studied French as a second language. The training consisted of three steps. The students were asked to provide their predictions on the text they were going to listen to after being told the topic, listened to the text, self-checked the

50

Chapter Two

accuracy of their predictions on their worksheet, and worked in pairs to compare and discuss what they had understood. Then, they listened to the text again and added more information to their worksheet. They were also allowed to discuss their strategies used for the listening comprehension in the class. A third listening allowed the students to reflect upon what they had learned about their own listening strategies and processes that worked for comprehension. Consistent with Nunan’s (1997) findings, the analysis suggested that the students perceived the training in a positive way, their motivation increased, and they understood their own thinking processes better. Ikeda and Takeuchi’s (2003) 8-week intervention involved 210 students of English in a Japanese university, investigating the impact of SBI on students’ use of some specific reading strategies (e.g., making inferences, using selective attention, using imagery, and summarizing). The participants were from two different proficiency levels (low & high). This study adopted a quasi-experimental design, dividing students at each proficiency level into an experimental and a control group. The post-test and survey of reading strategies indicated that change in frequency of strategy use only occurred to the high proficiency group. The explanation for this finding offered by the authors was that the strategies introduced only favored this group of students as these strategies mainly involve topdown processing, whereas the students in the low proficiency group might need bottom-up processing strategies. The delayed test showed that the high proficiency students retained their improvement in strategy use five months after the intervention. It was mentioned previously that speaking strategies were not included in the current review of LLS. However, as the design of Cohen’s (1998) intervention study of teaching students speaking strategies is of particular relevance to the present study, a review will inform the present study in the design. Cohen (1998) examined the effectiveness of SBI in universitylevel foreign language classrooms. The SBI intervention consisted of two major components: 1) showing students how, when and why strategies can be used for language learning through explicit instruction; and 2) embedding strategies into class materials through either explicit or implicit integration. All the participants (32 in the experimental group and 23 in the control group) were intermediate-level foreign language students at the University of Minnesota. A background questionnaire about the students’ previous language learning experience, reasons for studying the target language, contact with native speakers, visits to the target culture, current

Literature Review of Research on Language Learning Strategies

51

work schedule, grades in previous L2 courses, and college grade point average found no significant difference between the experimental and the control groups. The three experimental teachers received 30-hour training in SBI techniques for awareness raising, systematic strategy use in the classroom, SBI integration into lesson plans, and discussion of strategy effectiveness. The three control teachers received no training in SBI. The experimental students followed their normal syllabus as did the control students, but with strategies incorporated either explicitly or implicitly into their regular classroom activities during a 10-week period. The experimental teachers and students together generated a list of strategies useful for preparation, self-monitoring, and self-evaluation of three speaking tasks: selfdescription, story-telling, and city description. The experimental teachers’ retrospections about the structure and content of their classes were collected through individual and group interviews. Oxford’s (1990) SILL was administered both prior to and after the training of all students to determine the profiles and change of students’ LLS use. All the students were both pre- and post-tested on the same three tasks. To obtain data on the actual use of strategies, the experimental students were required to complete corresponding strategy checklists immediately after completing each of the three speaking tasks on the pre- and the posttest. The checklists were designed to capture their strategy use during the three stages of a speaking task: 1) preparation before the speaking task, 2) self-monitoring during the task, and 3) after-task self-reflection. For example, strategies such as rehearsal, note-taking, prediction of difficulties, attention to grammatical forms, and future learning plans were included. Four more questions were attached to the post-test version of the checklists for students’ self-reflection, three of which inquired about students’ experiences as language learners in relation to 1) the extent to which the tasks elicited their knowledge about the foreign language, 2) whether the tasks allowed for demonstration of the knowledge, and 3) their awareness level for language learning patterns and strategy use. The fourth question was designed to examine whether the students became more independent in language learning upon completing the 10-week strategy training. In addition, 21 students from both the experimental and the control groups representing three levels (high, medium, and low) of speaking proficiency were selected based on teacher ratings to provide verbal reports, as measures of their post-intervention SILL results in the form of think-aloud protocols. The checklists for strategies used before,

52

Chapter Two

during and after each task were specifically designed to make up the missing link between strategies and their use on specific language tasks, as a response to one of Gu’s (1996) comments on the common problems for most intervention studies in the literature. Like the majority of the intervention studies in the literature, the results of the study were mixed, though it adopted a more sophisticated design. The results of analysis of covariance showed that the experimental group did better than the control group only on the third speaking task in the posttest, after adjusting for pre-test differences. No significant difference was found for the first and second tasks. Both groups increased their use of some strategies included on the checklists. However, as noted by McDonough (1999), although the results of the study did not go all the experimental group’s way, there was a firm basis for the claim that SBI made a measurable difference in both how these students performed and in how well they performed, as measured by the strategy checklists, SILL, and verbal reports. Several limitations of the study have been pointed out by the researchers. First, the sample size of 55 subjects was quite small. Second, the experimental teachers were pursuing their PhD studies in education and applied linguistics, which supposedly helped them know more about language learning and teaching processes. However, the control teachers were all literature majors. Third, the control students may have benefited from being exposed to the SILL twice and the strategy checklists three times. The researchers have also noted many limitations and problems pertaining to their study. It is the researchers’ discussions of the limitations that deserve attention for this study. As a prevalent phenomenon, the majority of LLS research studies (both descriptive and intervention) avoid revealing their difficulties, limitations, and problems. As a result, the same problems continue to plague subsequent studies. Gunning and Oxford (2014) conducted an SBI on oral interactions among 54 year-6 ESL primary students in Quebec, Canada. The researchers adopted an experimental design with one experimental group and one control group. In order to achieve data triangulation, pre-tests, post-tests, children’s strategy questionnaire (Children’s SILL), video-recordings, observations, field notes, and a task-based strategy questionnaire were employed. The treatment students outperformed their control counterparts

Literature Review of Research on Language Learning Strategies

53

in their oral interactions at the post-test. In addition, the treatment group became more aware of their strategy use after the intervention. An interesting feature of the study is the school teachers conducted the SBI to their students, based on the lesson plans developed by the researchers and teacher collaboratively, which contributes to the utility and practicality of SBI (Gunning & Oxford, 2014). Such research might offer a practical working model to implement SBI in schools. Similar to intervention studies of reading, listening and oral interaction strategies reviewed above, only a small number of studies of writing strategies can be found in the literature. These writing intervention studies set out to measure change in writing strategy use (e.g., Lee, 2002), change in writing achievement or both (e.g., De Silva, 2015; Macaro, 2001; Sasaki, 2000). Some researchers have examined the effects of peer review/teacher feedback strategies, in particular, in writing instruction (e.g., Bitchener et al., 2005; Ferris & Roberts, 2001; Hu, 2005; Min, 2006). Donato and McCormick (1994) pointed out that discussions among peers are essential for learners to “share” strategies. According to socio-cultural views, social interactions, such as peer reviews, mediations and scaffoldings, occur in learning. Peer review strategies are particularly relevant to the present study in that peer review can be best perceived as socio-culturally situated and the present study will adopt a socio-cultural view as well. Sasaki (2000) compared writing strategies used by 12 Japanese EFL learners at different proficiency levels of writing (four professional writers, four more skilled writers, and four less skilled writers) and explored the effects of strategy-based writing instruction on eight writers (more skilled and less skilled) of the 12, using multiple data collection methods including essays, videotaped pausing behaviors while writing, and stimulated recall protocols after 6-month instruction. As mentioned by the researcher, a unique feature of the study was the collection of stimulated recall protocols from the student writers who watched their own videotaped writing processes right after the completion of writing. All the participants were asked to write an argumentative composition before the intervention. Only eight writers (four more skilled and four less skilled) completed the post-intervention composition. The composition results suggested that although the eight writers (also called novice writers in the study) scored higher on the post-test, the improvement was minimal. Their mean score was 130.71/200 for the preintervention task, whereas their mean score for the post-intervention task

54

Chapter Two

was only 136.5. The four professional writers only completed the preintervention writing task and achieved 190/200. Therefore, no cross-group comparison on the gains in scores between the two groups could be made. The qualitative analysis of the stimulated recall protocols revealed some differences in strategy use between the professional writers and the novice writers before the intervention. For example, the professional writers spent longer time in global planning (overall organization) than their novice counterparts. The novice writers paused more frequently as compared to the professional writers once they completed their global plan. The novice writers exhibited rigid global planning behaviors. For instance, once their plan was made, the novice writers would not adjust it when writing, whereas the professional writers could vary their plan based on their online assessment of the given task and elaborated but flexible goal-setting (Sasaki, 2000). This finding on writers’ planning strategies is in agreement with that of the studies of writing strategies (e.g., Gu et al., 2005; Raimes, 1985; Victori, 1999; Zamel, 1983) reviewed previously. After the intervention, a few global planning strategies and one re-reading strategy were adopted by the novice writers. They did not pause when writing as often as before. The eight novices spent more time before starting to write than prior to the intervention. The findings of this study suggested global planning strategies are of importance with regard to the difference between the skilled and less skilled writers. The global planning strategies exhibited by the expert writers could be taught to the less skilled ones through intervention. Macaro’s (2001) study involved six classes of secondary students of French which were randomly assigned to an experimental group and a control group. An experimental design was adopted that was comprised of both pre- and post-tests, a strategy questionnaire, writing tasks, and thinkaloud interviews (a sample of 18 students) during a French writing task. The experimental students received strategy instruction on a variety of strategies that included such metacognitive strategies as advance preparation, monitoring, and evaluating over a period of five months. By describing two case studies, Macaro demonstrated two types of change: 1) learners’ change in their use of writing strategies as a result of instruction in writing strategies for five months, and 2) their change in writing processes (planning, composing, and checking). The training included: 1) making the right decision of when to use the L1 and when to use the L2 in planning and composing; 2) dictionary strategy use; and 3) self-monitoring strategies. Significant changes in terms of language competence and strategy use were noted in the experimental group. Upon completion of the

Literature Review of Research on Language Learning Strategies

55

intervention, the students in the experimental group committed significantly fewer grammatical errors, as compared to their counterparts in the control group. Besides, the experimental students also reported change in their approach to writing as a result of the training. For example, they became less dependent on teacher materials and the dictionary. The questionnaire results indicated that before the intervention, the experimental students were “higher dictionary users”, as compared to the control students. However, they appeared to be more selective in dictionary use and more careful about their writing after the training. This study exhibited the following strengths that can be of particular use to the present study. First, the two case studies provided more comprehensive insights into writers’ writing strategies and processes as well as changes in strategy use and writing processes over the intervention period than a strategy questionnaire. Second, a writing questionnaire was designed specifically for the study. This suits the participants’ profile of strategy use better than simply adopting a writing strategy questionnaire developed by other researchers (Nisbet et al., 2005). As noted by Chamot (2004), taskspecific questionnaires may better serve the purpose. Third, the intervention results were analyzed based on data from multiple data sources: 1) the test results and the self-reported strategies associated with the test, 2) the error analysis of the task and the associated self-reported strategies, 3) the think-aloud transcriptions and the corresponding task, 4) questionnaire findings, and 5) the teachers’ report of students’ strategy use. This served the purpose of triangulation very well and is particularly of relevance to the present study. In particular, Macaro identified a group of 18 writing strategies which could help writers make better decisions in the process of writing. He suggested that “…researchers have to find out from the particular groups being researched their current strategy rather than relying on previous research and then embarking on some form of learner training” (p. 134). As reviewed previously, most research studies just simply adopted strategy questionnaires with the assumption that their participants used the same strategies as those from other studies. The strengths notwithstanding, the study also suffered from a few shortcomings. The detailed description of two case studies may not be very representative though providing insightful information on the two students’ strategy use and the change that the intervention gave rise to. More case studies of students in the experimental group could have been analyzed for the purpose of generalization. Second, both cases were extracted from the experimental group, which does not allow for a

56

Chapter Two

comparison in the change of strategy use between the experimental and the control students. The two cases might have received other treatment that helped improve their writing during the intervention period. In order to rule out the possibility that the improvement was not a result of factors other than the writing intervention, a few cases from the control group could have served the purpose. Lee’s study (2002) examined the effects of strategy and self-regulation instruction on 29 Malaysian ESL learners who were university engineering students. The data for this 7-week study were collected through a pre- and post-intervention questionnaire on writing processes, a self-regulation questionnaire, teacher’s observations, and students’ essays. The SBI consisted of three stages: pre-session, in-session, and post-session. At the pre-session stage, a questionnaire on holistic writing processes was administered to the students. The strategies taught at the in-session stage were mainly planning and revising strategies. The self-regulation questionnaire-based survey was conducted after each of the five SBI sessions to record the writing strategies employed by the students when writing. The teacher also observed the students on their strategy use during discussion sessions. At the post-session stage, the same questionnaire (writing processes) was administered. The students’ essays written during the intervention were analyzed qualitatively by the teacher (the researcher) to see whether they had made improvements on strategy use, as compared to their essays done before the strategy instruction. As reported by Lee, no significant differences were found between the two questionnaire (pre- and post-intervention) administrations. However, analysis of data collected with the five administrations of the selfregulation questionnaire indicated that the students made significant improvements on their planning and revising strategies as a result of the utilization of four self-regulation strategies: self-evaluation, organizing and transforming, seeking information, and seeking assistance. After the intervention, the students became more positive towards negative peer feedback and realized that writing more without analysis of their own mistakes did not help improve writing. Although a few data collection methods (e.g., questionnaires, teacher’s observations, and students’ essays) were used in the study for the purpose of data triangulation, no comparisons between the teacher’s observations and findings from the questionnaires were reported. Therefore, the findings of the study should be interpreted as tentative. Furthermore, that the same self-regulation questionnaire was administered five times is a legitimate cause for concern

Literature Review of Research on Language Learning Strategies

57

in terms of testing effects. It can be argued that the findings may have been affected by the students’ perceptions of the researcher’s aim as a result of their completing the questionnaire five times and a tendency to please their teacher. De Silva (2015) undertook an intervention study on writing strategy use and writing achievements of university students in Sri Lanka. This experimental study used a pre- and post-test design including the pre/post strategy questionnaire and a writing task battery. An important feature of this study was the full cycle of writing strategy instruction (e.g., from goal setting to evaluation), which is typical of writing instruction of the real classroom. Overall, the experimental group showed greater increase in their strategy use than the control group. In particular, the experimental group outperformed their counterparts in their use of planning, formulating, resourcing, monitoring, evaluation and revision strategies. In terms of writing achievements, the experimental group also made significantly greater improvements than the control group. The findings suggest that the SBI exerted a positive impact on both strategy use and writing performance of the experimental group. One useful feature of this study is the inclusion of all the writing stages that student writers are engaged in for writing, which is particularly relevant to the present study. The present study aims to measure the effects of writing strategies used by Singapore primary pupils use in the whole writing process for completing a composition. In light of the recent development in writing instruction, “peer review has a prominent place in process-oriented writing instruction” (Hu, 2005, p.321). Peer review entails collaboration among learners who read, critique, and provide feedback on each other’s work. It can be a very useful writing strategy. Potential pedagogical benefits, such as textual improvement and increasing writing competence via mutual scaffolding (Vygotsky, 1978) have been noted by many researchers (e.g., Ferris & Roberts, 2001; Ferris & Hedgcock, 2005; Hu, 2005). Furthermore, this line of research shares the view that learners benefit from social mediation (see chapter 3 for a detailed discussion of mediation) with their co-learners. In the classroom, students learn together most of the time. Therefore, students’ response to peer review is another area of writing strategy use that is important to explore, especially from a socio-cultural perspective. A study on peer review strategies was conducted by Hu (2005) taking the form of action research. The participants were all Chinese nationals

58

Chapter Two

studying in an English preparatory course in Singapore before their entry to university. The training comprised two major components: awareness raising and explanation of procedures. Though well structured, Hu’s training in the students’ use of peer review in the first year of the action research project failed to produce positive results. He attributed the unsuccessful attempt to the students’ inadequate understanding of the peer review processes, the training, and the potential benefits of peer review. In his subsequent attempts, he refined the training process. For example, awareness raising included more well-organized steps, e.g., small-group discussions of benefits, problems and solutions, teacher summary of research findings on peer review, and real examples of writers benefiting from peer review, as compared to only whole-class discussion about peer review in the previous training. In addition, more components (i.e., demonstration, practice, reflection, and pre-response review), each with detailed steps, were incorporated to ensure effects. As a result, Hu (2005) found the students developed much better attitudes towards peer review. Min (2006) worked with 18 first-year university students in Taiwan to train them on peer review in two major phases: in-class modeling and oneon-one conferencing. For modeling, the instructor firstly demonstrated how peer review worked through her think-alouds in four steps: clarifying writers’ intentions, identifying the source of problems, explaining the nature of the problems, and making specific suggestions with examples. The students were then asked to review their friends’ writing according to the probing questions given. In one-to-one conferencing, the instructor discussed with the reviewer his/her comments on the writers’ drafts and revisions so that the reviewer could improve on his/her comments. The findings suggested that the peer review instruction had a positive impact on the student reviewers’ ability in providing comments, the communication among the student writers and the reviewers, and their subsequent revisions (cf. Hu, 2005). The results indicated that 77% of the instructed peer review feedback was incorporated into the students’ subsequent revisions. Similar to Hu’s (2005) research findings, the student writers found peer review useful and were willing to apply it. The above-reviewed studies of peer review have important implications for the present study. Though the studies were in the tradition of processoriented writing instruction, they brought to attention the understanding that peer review is a socially constructed process. During the process of peer review, writers engage themselves in collaborative learning in which they also receive social support and scaffolding from their peers. This is a

Literature Review of Research on Language Learning Strategies

59

good example of how social-cultural dimensions of literacy can be woven into writing strategy instruction, an area that has been hitherto dominated by (cognitive) process-oriented approaches.

2.3.3 Critique of intervention research on LLS The above reviewed intervention studies (general and specific) do not warrant firm conclusions about the relationships between strategy use and actual language performance, as noted by Gu (1996). Most intervention studies only showed increase in learners’ frequency of LLS use, partly due to the nature of intervention research. Rigorous intervention research requires at least the following two criteria. First, a rigorous intervention study should have randomly formed experimental and control groups. In reality, however, it is often impossible to achieve that in a natural school setting. Most of the aforementioned intervention studies (e.g., Cohen, 1998; Ikeda & Takeuchi, 2003; Kuba, 2002; O’Malley et al., 1985b; O’Malley & Chamot, 1990) were only able to adopt quasi-experimental designs. Some (e.g., Lee, 2002; Min, 2006) used non-experimental designs due to the constraints in the school setting. Second, instruction in each group should be identical except for the presence or absence of the strategy part. Preand post-tests are an integral part of any intervention study to gauge the differences made. Other factors, such as motivation and attitude, must also be taken into consideration. However, the above-mentioned criteria for a successful intervention study are difficult to meet in a natural classroom setting. Nonetheless, the review of LLS intervention studies has identified several significant implications for the present study. First, as teachers play an important role in classroom instruction, teacher training in SBI should be provided as a necessary component in an intervention study if it is to have ecological validity and generalizability as well (Gu, 1996). The findings of Gunning and Oxford (2014) also suggest that teachers’ training and involvement in SBI is critical in helping students learn. However, in many previous research studies (e.g., Carrier, 2003; De Silva, 2015; Lee, 2000; Sasaki, 2002) investigating LLS, the researchers served as LLS instructors in the classroom. Less attention has been paid to teacher training in LLS research. In light of such findings, classroom teachers should conduct SBI in their own classrooms, with researchers’ assistance in future research. Second, it is highly necessary for future research to incorporate a more rigorous methodology, e.g., pre-, post and delayed tests, and experimental

60

Chapter Two

and control groups to measure the effects of strategy training more rigorously (Chamot, 2005). As reviewed above, the results of LLS intervention studies tend to be mixed, partly due to the varying research methods that these studies adopted. For example, Lee’s (2002) study did not include a control group, which undermined her claim about the effectiveness of her training. Specifically, the delayed test represents an effort to find out if the obtained effects of intervention sustain. Although the various constraints inherent in the Singapore school context prevent the present study from adopting a genuine experimental design that involves randomly formed experimental and control groups, a quasiexperimental design can serve the aims of the present study adequately. A third implication is the generalizability of intervention research on LLS. The majority of the intervention studies reviewed previously involved a small number of participants. For example, only seven participants were involved in Carrier (2003). The research findings from small-scale studies (e.g., Carrell et al., 1989; Lee, 2002; Min, 2006) are less generalizable. However, these studies adopted a qualitative approach. Although qualitative research allows for in-depth analysis of LLS use of individual learners, quantitative data, such as results of writing tests and questionnaires, also constitute a necessary component for evaluating the effectiveness of strategy intervention. Either evaluation method only provides a partial view of the effectiveness of the intervention (Chen, 2007). Therefore, the present study may be better off employing both quantitative and qualitative research methods. Fourth, explicitness of strategy instruction is also an important issue to take into account. Two strands of SBI, i.e., implicit strategy instruction and explicit strategy instruction, coexist in the literature. In implicit strategy instruction, the instructor guides learners through activities with strategies embedded in the activities, but does not inform them that they are actually employing strategies. By contrast, explicit instruction requires the instructor to inform learners of the benefits of strategy use and demonstrate how to apply and transfer the strategies to new contexts or tasks. “Explicit instruction includes the development of students’ awareness of their strategies, teacher modeling of strategic thinking, identifying the strategies by name, providing opportunities for practice and selfevaluation” (Chamot, 2005, p. 123). Chamot (2004, 2005) suggests that implicit strategy instruction does not facilitate maintenance and transfer of strategies, but explicit instruction leads to maintenance and transfer of strategies. Researchers (e.g., Carrier, 2003; Duffy, 2002; Grenfell & Harris,

Literature Review of Research on Language Learning Strategies

61

1999; O’Malley & Chamot, 1990) have provided evidence for the effectiveness of explicit strategy instruction. Anderson (2005) also supports the use of explicit strategy instruction through reviewing a series of LLS studies that showed positive results using this method.

2.4 Conclusion In this chapter, I firstly reviewed definitions and classifications of language learning strategies. Varying definitions and classifications can be found in the literature due to researchers’ differing theoretical frameworks within which their LLS research was carried out. Although consensus on these issues is yet to be reached, our understanding of LLS has been much improved because of the research studies conducted. O’Malley and Chamot’s (1990) definition and classification seem to be the best fit for the present study. Although many descriptive LLS studies have found that successful language learners use more LLS than their less successful counterparts, the relationship between strategy use and language learning outcome presents a complex picture for many other studies. With regard to LLS intervention research, it is also difficult to generalize the findings of the previous research as those studies used very different research methods and designs. Lack of theoretical underpinnings as a problem may still persist in the LLS research. However, researchers (e.g., Donato & McCormick, 1994; Ferris & Hedgcock, 2005; Hu, 2005; Min, 2006) have begun to explore and look at language learning strategies from wider theoretical perspectives. For example, Donato and McCormick (1994) suggest that socio-cultural theory is a robust framework for LLS research. Atkinson (2003b) notes that writing is not only cognitively process-oriented but also socioculturally based. The multi-dimensional view on literacy (Kucer, 2005, 2009) also suggests that reading and writing, two aspects of literacy, should also be examined from a socio-cultural perspective in addition to its inner aspect, the cognitive dimension. According to Kucer’s framework, language learning strategies are mechanisms that help learners with their literacy learning. In chapter 3, I will propose my theoretical framework, which examines language learning strategies from multiple perspectives. To conclude, the relationship between strategy use and proficiency is manifested with complications. A simple linear relationship between LLS use and language proficiency could not be established in LLS research. As

62

Chapter Two

a matter of fact, if we begin to look at language learning strategies as socio-culturally constructed phenomena, we have to expect more complex relationships between strategy use and proficiency to be revealed in LLS research since not all writers go through the same writing processes and at the same cognitive developmental stage. As regards LLS training, despite the extensive work done in other countries, very little is known about LLS use and effects of LLS training on Singaporean bilingual children. The literature review in this chapter shows that learners from different cultures, and educational settings have exhibited different choices of LLS. Therefore, there may be distinctive features of strategy use by Singaporean bilingual children, as compared to those in other cultures.

CHAPTER THREE THEORETICAL FRAMEWORK FOR THE PRESENT STUDY

The literature review in the previous chapter identified the need for studying language learning strategies in relation to learners’ achievements and their literacy development as an overall goal. However, there is no comprehensive and consistent theoretical framework to inform LLS research. As shown in chapter 2, the majority of the LLS studies conducted were theoretically grounded in cognitive psychology (Macaro, 2006). Works by Oxford and O’Malley and Chamot are such examples. From this theoretical perspective, the actual teaching of strategies in the classroom mainly focus on the cognitive processes or cognitive development of students. However, as has been proposed by Kucer (2005, 2009), literacy has multiple dimensions, i.e., cognitive, linguistic, sociocultural and developmental. The multi-dimensional view can widen our understanding of SBI in particular and literacy in general. Therefore, building on previous theories of LLS, this chapter will propose a new theoretical framework for looking at LLS and SBI in relation to a multidimensional view on literacy (Kern, 2000; Kucer & Silva, 2006). Section 3.1 will examine learning strategies informed by cognitive psychology. The focus is on the active role of the individual learner in meaning making and their differences in strategy use. Specifically, Anderson’s Adaptive Control of Thought (e.g., 1976, 1982, 1983, 1985, 1992, 1993) will be analyzed for LLS from a cognitive perspective. Then, the socio-cultural dimension of literacy will be discussed in section 3.2 as the acquisition of LLS can be explained in this dimension. Kucer (2005, 2009) notes that literacy events are much more than the cognitive processes that individual learners go through for meaning making. Literacy is also a social act. This dimension of literacy suggests that, whereas cognitive views on learning strategies are useful to account for individual differences in learning, socio-cultural theories (e.g., Vygotsky, 1962, 1978; Feuerstein, 1990) can explain how individuals learn through

64

Chapter Three

their social interactions with each other. According to socio-cultural theories, social and individual processes in learning depend on each other. The literacy development of learners mainly takes place through interactions with others (Coyle, 2007). Therefore, SBI can be better theoretically grounded in this regard. After the discussion of the cognitive and socio-cultural perspectives on literacy in relation to language learning strategies and LLS intervention, section 3.3 will discuss a multiperspective view on literacy and LLS (Kern, 2000; Kucer, 2005, 2009; Kucer & Silva, 2006). Lastly, I will propose an integrated model in which LLS and SBI are situated in multiple dimensions of literacy.

3.1 Cognitive views on language learning As mentioned in chapter 2, researchers have not reached a consensus on the definition of language learning strategies as a result of atheoreticality among them. For example, Oxford (1989) perceives LLS as behavioral while Weinstein and Mayer (1986), Chamot (1987), and O’Malley and Chamot (1990) regard them as both behavioral and mental. No consensus on the theoretical foundation of LLS was reached to investigate the nature of language learning strategies and their influence in learning an L2 at the onset of LLS research. Researchers, such as Rubin (1981), were primarily concerned with finding out the differences between successful learners and their unsuccessful counterparts in their approaches to learning an L2. The majority of those early LLS studies were descriptive, investigating the LLS used by effective learners, i.e., the “good language learner”. The research field of LLS suffered from a lack of grounding in comprehensive theories to describe what learning strategies are, what role LLS play in language learning as well as a framework for successful intervention. In subsequent LLS studies, cognitive theories were employed in LLS research work. Researchers (e.g., Rabinowitz & Chi, 1987; O’Malley & Chamot, 1990; O’Malley, Chamot, & Walker, 1987; Wenden, 1987) were more concerned with explaining LLS from a theoretical perspective, besides accounting for learners’ activities that made them better learners. For example, Weinstein and Mayer (1986) viewed LLS as cognitively oriented. They defined learning strategies as “behaviors and thoughts that a learner engages in during learning that are intended to influence the learner’s encoding process” (1986, p. 315). O’Malley and Chamot (1990) also based their research work on a cognitive view of skill acquisition, i.e., Anderson’s (1983) Adaptive Control of Thought (ACT). To them, learning strategies are basically cognitive oriented, similar to cognitive skills. The

Theoretical Framework for the Present Study

65

basic concepts involved in ACT are declarative knowledge, procedural knowledge, and production systems. Two important aspects of cognitive views on skill acquisition with regard to human thoughts and actions can be concluded: (1) human behaviors can be best understood by reference to how individuals perceive and interpret their experiences, and (2) there exist parallels between how the computer processes information and how a human being thinks and reasons (e.g., Shuell, 1986; Johnson, 1996; McLaughlin, 1987). In this light, learners are perceived as organizers of information, with information processing capacity like a computer. The thoughts involved in processing are referred to as “mental processes” or “cognitive processes”, and learning strategies are special ways of information processing that individuals are engaged in to improve comprehension, learning, information retention, and the like (O’Malley & Chamot, 1990).

3.1.1 Anderson’s theory and language learning It has been mentioned in the previous section that Anderson’s ACT (1976, 1983, 1993, 1995) has been widely applied to LLS research. According to ACT, language learning can be regarded as cognitive skill learning. There are several advantages to this view. First, the development in cognitive psychology (Anderson, 1983, 1985, 2000) and computer information processing (Gagné, 1985; McLaughlin, 1987, 1990) has advanced our understanding of cognitive skill acquisition. For example, Anderson’s (1982, 1983, 1995) ACT describes the configuration of learning mechanisms involved in the full range of skill acquisition from language acquisition to problem solving with concrete examples. Second, viewing language learning as cognitive skill learning situates language learning in the context of a dynamic process. Language learning is a process in which learning strategies are repeatedly modified and fine-tuned. Third, LLS, referred to as cognitive processes in cognitive psychology, can be perceived as mechanisms that can enhance language learning. In this section, I will discuss Anderson’s ACT in detail to guide the present study from a cognitive perspective. A very important concept in Anderson’s ACT is production systems (Anderson, 1982, 1983, 1985, 1987, 1995, 2000), which has provided a theoretical basis to view language learning as cognitive skill learning. Anderson (1983) called for a unitary theory of the mind or a common cognitive system to understand and explain how complex cognitive skills

66

Chapter Three

are represented in human memory. The production systems show us a model of how complex cognitive skills can be represented in memory by means of uniform principles of representation. He explains that all complex cognitive skills can be represented as production systems. In addition, language can best be understood as a complex cognitive skill, and the mental processes for language learning parallel those for other cognitive skills both in memory representations and in learning. Therefore, the process of learning a language, be it a first or second language, is essentially the same as any other kind of learning, cognitively. A production system is a model of cognitive processing, consisting of a collection of rules, called production rules (systems), or simply productions. It is these production systems that provide the basis for a unitary theory of cognition. Each production can be seen as a rule, which has a “condition” and “action”, taking the form of IF-THEN clauses used in computer programming. The condition may contain a clause or clauses proceeded by IF, and the action has a clause or clauses proceeded by THEN. When the condition holds true, the action is taken. For example, Anderson (1980, p. 239) demonstrates the following production for “pluralization”: IF the goal is to generate a plural of a noun and the Noun ends in a heard consonant, THEN generate the Noun + s

The productions or rules are the basis for language production, which range from plural productions to SVO order (Anderson, 1983, p. 262) (LV stands for Local Variable). IF the goal is to communicate a meaning structure of the form (LV relation LV agent LV object), THEN set as sub-goals 1. to describe LV relation 2. to describe LV agent 3. to describe LV object

The production systems work in a recognize-act cycle manner. First, a rule whose conditions are satisfied is identified. Then, the rule is triggered; its action or actions are carried out. The actions may change the current situation so that the conditions for a different rule are satisfied, and the cycle repeats. The above example of production systems can better demonstrate how such systems work: IF you want to say something about an agent being related to an object, THEN you need to (1) describe the agent; (2) describe the relationship; and (3) describe the object. The rule

Theoretical Framework for the Present Study

67

generates three new sub-goals (actions), which will be achieved (carried out) in turn by other production rules, thus resulting in another recognizeact cycle or cycles. At the initial period of learning, the recognize-act cycles can be stored as declarative knowledge in memory and then through repeated practice, be proceduralized. Sequences of steps of learning are integrated into one action. In this regard, the production systems demonstrate how language learning works in a dynamic way. Some important terms need to be discussed in order to address how individuals acquire new information and how information is stored and transformed in memory. Memory can be simply categorized into two forms: short-term memory (or working memory), which stores a limited amount of information for a limited amount of time, and long-term memory, which can hold information for lengthy periods of time and may be represented as isolated elements or as interconnected networks. Two memory stores (declarative & procedural) exist in the long-term memory (see Figure 3.1). Anderson’s production systems (1983, 1995) draw a distinction between declarative knowledge and procedural knowledge. Declarative knowledge represents what we know about or static information in memory, whereas the knowledge of how to do things or dynamic information is stored in the procedural form, and is called procedural knowledge. Examples of declarative knowledge include, but are not limited to, word definitions, facts, rules, events, and images. This type of knowledge is represented in terms of chunks, which are structured as a semantic network, consisting of nodes (ideas) and links (associations). Procedural knowledge constitutes our ability to understand and create language or solve problems using our knowledge of rules. This type of knowledge is represented as production rules in the form of IF-THEN (conditions-actions) pairs that were discussed previously, which are the basis for explaining how complex cognitive skills such as language are learned and used. According to Anderson, knowledge usually starts in the declarative form (Anderson and Lebriere, 1998) and leads to the procedural form. The move from declarative to procedural knowledge goes through three major stages: declarative stage, associative stage, and procedural stage (Anderson, 1983, 1985, 2000; Anderson & Lebriere, 1998). The acquisition of a cognitive skill, such as language, is a progressive and dynamic process that progresses from rule-bound declarative knowledge to procedural knowledge. The learning process starts with the interpretative application of declarative knowledge in the declarative stage. Then, it

68

Chapter Three

proceeds to compile declarative knowledge into production rules during the associative stage. Gradually, the production, or a collection of rules, becomes increasingly fine-tuned, and autonomous with extensive opportunities for practice.

Figure 3.1. A General Framework of Anderson’s ACT Production System (1983, p. 19)

At the beginning of the process of skill acquisition (the declarative stage), new information enters memory in declarative forms. During this stage, individuals learn a collection of facts relevant to the skills, such as facts or descriptions of images and events. Before a learner can speak a second language fluently, the learner may try to memorize some words and the grammar rules to produce utterances. For example, the learner may know that “worked” can break into “work” and “-ed”, while s/he may not be able to produce it correctly in a conversation. The learner may also memorize chunks of language and facts about the language. The processing at this stage is slow, conscious, and deliberate, and needs full attention. Only with conscious attention and repeated practice, may the learner proceed to the next stage.

Theoretical Framework for the Present Study

69

The associative stage is also known as the compilation stage; knowledge compilation occurs at this stage. This stage is to compile the information into more specific procedures and to produce procedures which can be executed more quickly and appropriately. One form of knowledge compilation is realized through “composition”, by which several productions are merged into one. Errors in the original declarative form are gradually detected and deleted through practice. This is the stage where declarative knowledge is converted into procedural knowledge; the connections among the various elements or components of the skill are strengthened (O’Malley & Chamot, 1990, p. 26). For example, the leaner may have learned “worked”, “stopped”, and “pointed” separately and come to realize a more efficient and economic production set is possible. In this case, “-ed” can be added to the verb to form a past tense verb. An important result of this stage is that the working load for the short-term memory is much reduced and therefore the learner can achieve some efficiency in learning and production of language. At the autonomous stage, the procedure becomes increasingly automated and rapid. The most salient difference between this stage and the associative stage is that productions are more fine-tuned, to the point where errors and ill-productions are greatly reduced, and even disappear. The right productions (rules) can be easily located and executed without much effort (Anderson & Lebiere, 1998). Three basic mechanisms take part and serve as support for this stage: generalization, discrimination, and strengthening, which can also be seen as three cognitive processes. The generalization process is for learners to extract commonalities from different productions so that broader production rules in terms of applicability can be created, whereas the discrimination process is to narrow down the production rules so as to restrict the ranges of application of the rules to the most appropriate circumstances. In this stage, learners are also making improvements in their selection of production rules for the appropriate contexts, thus strengthening their degree of autonomy. For example, learners may start to realize that the past tense verb production set mentioned above is only applicable to one set of verbs. Verbs such as “fly” and “write” fall out of this production set. Learners may reformulate the production set to accommodate a wider range of verbs as exemplified above. “The ACT model is enormously complex” (Ellis, 2008, p.429). Learning an L2 as described above is analogous to acquiring a cognitive process or skill. It does not only “involve the speeding up of the same procedures

70

Chapter Three

formed originally from declarative knowledge, but also the establishment of new procedures reorganizing previously acquired procedures” (McLaughlin & Heredia, 1996). To put it simply, ACT explains that learning (e.g., L2) normally begins with declarative knowledge and transits into procedural knowledge. As asserted by O’Malley and Chamot (1990), ACT, among other existing cognitive theories, can better explain how language and cognition interact and why language acquisition parallels the way complex cognitive skill learning occurs. Therefore, language learning has been characterized as special cognitive processes learning, thus allowing us to perceive the mechanisms for enhancing language learning, namely LLS, as cognitive processes that make learning more effective. A detailed discussion of LLS as special cognitive processes follows.

3.1.2 LLS as special cognitive learning processes As reviewed in the previous section, Anderson’s (1980, 1983, 1985, 1996, 2000) cognitive psychology, which were originally designed to theorize on how information is stored and retrieved, has influenced researchers of language learning in general and LLS researchers (e.g., O’Malley & Chamot, 1990; Oxford, 1999; Weinstein & Mayer, 1986) in particular. It also informs the theoretical framework of the present study from a cognitive perspective. While Anderson’s theory describes how information is processed rather than how learning can be improved, O’Malley and Chamot (1990), influenced by Anderson’s theory, view LLS as complex cognitive processes and as component mental processes that can be taught to learners as special ways to enhance learning. Their findings show how cognitive theory can be applied to the acquisition and use of LLS. In their view, strategies can be represented as mental processes and described as a collection of rules for proceduralization. To further understand LLS in the context of Anderson’s production systems, each strategy can be perceived as configuration of a series of necessary actions governed by a set of rules as mentioned in the foregoing section (3.1.1). For example, in order to understand the meaning of a word in a reading passage, learners may take, but are not limited to, the following actions. First, they may try to recall if they have seen the word elsewhere and then guess the meaning. Second, they may infer the meaning from the local context. Third, they can consult a dictionary. Fourth, they may decide to continue reading and later check if they can understand it. Fifth, they may ask other people for help. There are rules for

Theoretical Framework for the Present Study

71

using these actions. If learners think that it is important to understand the meaning, they will make sure that they will find out the correct meaning immediately. They may use one of the first three actions. If they do not think that the word will affect their comprehension, they may employ action 4. If they still cannot figure out the meaning, the fifth action may be undertaken. At the declarative stage, the actions firstly exist in the form of declarative knowledge. Learners may have to deliberately assemble or recall these actions while figuring out the best rules for these actions in order to use/try out a strategy. At this stage, when encountering a new word in a reading passage, learners may only look the word up in the dictionary to solve this problem. They may go through all the entries for that word and then decide the meaning. In fact, they may not have a problem in comprehending the text even without knowing the meaning of the word. They can guess the meaning later when they finish reading the passage if the word does not affect their understanding. However, they will always look up the meaning of the word that they do not know in the dictionary. Therefore, the deployment of a strategy may be fraught with errors since the rules have not been fine-tuned yet. In other words, learners still cannot recognize the application of the actions that are governed by rules (Anderson, 2005). Only with more practice, are learners able to employ the strategy more successfully. At the associative stage, learners can easily recognize the actions with appropriate rules. As was discussed previously, declarative knowledge is converted into procedural knowledge at this stage. In other words, the rules or the connections among the actions of a strategy are more strengthened (O’Malley & Chamot, 1990). For the same example mentioned above, learners may have developed some strategies in order to understand a new word at this stage. They are aware of the actions available, which are rule-bound. For example, they may use one or some of the actions whenever they encounter a new word. Depending on the difficulty level of the vocabulary, they understand that they do not need to stick to the same action only for tackling such a learning problem. At this stage, learners begin to realize that the appropriate actions should be used when the underlying conditions or rules are met. At the final stage of proceduralization, learners are more proficient with their employment of LLS. A few production rules may collapse into one production rule, which may later serve as one of the production rules for

72

Chapter Three

another bigger production rule. For example, instead of attempting the five actions one by one, they may first assess the extent to which that the meaning may affect their understanding. If the meaning is highly important, they may consult a dictionary directly in order to know the accurate meaning of that word. Otherwise, they may just infer the meaning from the context or take another action. At this stage, learners may not think deliberately what strategy or strategies to use for the language problem at hand, they can easily activate one strategy (or a cluster of strategies) that is most appropriate for the problem (Anderson, 2005; Anderson and Lebiere, 1998). It is important to point out when learners use their strategies without effort and automatically, their LLS have become learning skills (Afflerbach et al., 2008). O’Malley and Chamot (1990) draw a distinction between two types of processes in Anderson’s theory. The two types of processes are metacognitive and cognitive processes. Metacognitive processes involve thinking about language or the learning process, including such processes as planning, monitoring, and selective attention (Grenfell & Harris, 1999). Cognitive processes involve working directly on and manipulating the incoming information, including such processes as rehearsal, organization, and inferencing (Weinstein & Mayer, 1986). LLS can be seen as a subset of metacognitive and cognitive processes used by the learner that bear the power for learning improvement. O’Malley & Chamot (1990) added a third group of learning strategies, namely social/affective strategies. They are “a broad group that involves another person or ideational control over affect” (p. 45), including strategies like cooperation, questioning for clarification, and self-talk. A few metacognitive processes were identified as effective in improving learning: planning, monitoring, and selective attention in Anderson’s theory. These metacognitive processes, which may exist as procedural knowledge in production systems for examining, testing and modifying the procedural system as well as extending the system’s range of control, are the metacognitive strategies identified in LLS research. These processes, to a great extent, are identical or similar to the metacognitive strategies described by O’Malley and Chamot (1990). Planning, according to Anderson (1983), is to resolve the conflicts among the competing action statements. This step applies to the conditional (IF) clause in the production systems, and may be influenced by goals, which can be further divided into higher-level goals and lower-level goals.

Theoretical Framework for the Present Study

73

Learners can be viewed as top-down processors when successfully refining higher-level goals into achievable actions. If they start with features of input or lower-level goals, they can be called bottom-up processors. Learners may alternate their processes depending on the difficulty of their learning tasks at hand. O’Malley and Chamot (1990) capitalized on this view in their LLS research. For example, they found learners may alternate processes depending on the difficulty of the specific language items contained in the text, or at which stage of learning learners happen to be functioning with any item of input. In their listening comprehension research, they noted that listeners may employ a top-down approach to listening by attending to the overall meaning of phrases and sentences. Similar findings were also reported by Vandergrift (2003a). However, the difficulty of the listening tasks may require listeners to attend to specific linguistic features, which a bottom-up approach entails. The IF clause(s) in the production systems can be perceived as a goal statement, thus providing direction for planning future actions, the THEN clause(s), since learning is a goal-driven action, cognitively. O’Malley and Chamot (1990, p. 52) provide the following examples as strategy application illustrations with a condition (IF) and one or more action (THEN) clauses: IF the goal is to comprehend an oral or written text, and I am unable to identify a word’s meaning, THEN I will try to infer the meaning from the context. IF the goal is to comprehend a concept in a written text, and I know the concept is not at the beginning, THEN I will scan through the text to locate the concept.

As reviewed in chapter 2, many LLS researchers (e.g., Gu et al., 2005; O’Malley & Chamot, 1990; O’Malley et al., 1985a; Stern, 1975; Vandergrift, 2003a; Victori, 1999; Zamel, 1983) also noted the use of planning by effective learners. For example, Gu et al.’s (2005) effective writers frequently used planning strategies in their writing process. Monitoring is another important mental process, which is relevant to the study of LLS. “Monitoring is a response to ambiguity in comprehending language where a learner selects a best guess of the meaning based on available information” (O’Malley & Chamot, 1990, p. 48). For example, when writing, the writer may remind him/herself of sticking to the plan. The writer may also look for and attend to mistakes, be it grammar or spelling while writing. The importance of monitoring has been reflected in

74

Chapter Three

LLS research. For example, good learners are aware of the monitoring process and employ the process, which can distinguish themselves from poor learners (e.g., Gu et al., 2005; Nisbet & Shucksmith, 1986; Nisbet et al., 2005; Rubin, 1981; Vandergrift, 2003a). Selective attention is also an important metacognitive process, which occurs during listening comprehension in Anderson’s (1983) theory. According to information processing theory, human beings are limited in both their capacity and scope for information processing. However, the tasks involved in a communication process may be very complex and difficult. Not everything that reaches the organism of human beings through various input channels can become an object of attention automatically. Therefore, selective attention is employed to process information. When individual learners deem certain aspects of the information less important, they will attend to the more important information selectively in order to reduce the cognitive load that the new information imposes on them. In other words, highly strategic learners can attend to things more selectively and process information more automatically, as compared to those less strategic ones who need to pay their full attention and process the information in a controlled manner. O’Malley and Chamot (1990) observed that L2 listeners may be unable to tell the word boundaries from phrase boundaries at the beginning of their learning. Their failure in identifying meaningful chunks or segments of text resulted in their inability to attend to the text appropriately for successful comprehension. Thus, listeners should be trained to use such strategies as attending to key words, phrases, or word/phrase boundaries. O’Malley et al.’s (1985b) study incorporated selective attention as a component into their intervention. Before listening to a text, the students were told or reminded of what and how to listen. Key words (linguistic markers) or specific items, such as nouns, and the answers to the comprehension questions were their focus in listening. Students were also reminded to ignore irrelevant background noises while listening. As a result, the analysis showed that the students made progress on the tasks which were more interesting and less demanding. The researchers, therefore, suggested that strategy training can be effective, depending on the difficulty of the materials. The SBI research done by Carrier (2003) and Ikeda and Takeuchi (2003) also trained learners in using this strategy. Planning, monitoring, and selective attention have been seen as important metacognitive processes of individual learners both in cognitive psychology and in LLS research. LLS researchers, who have been

Theoretical Framework for the Present Study

75

influenced by cognitive psychology, have been looking at the metacognitive strategies that good language learners use for learning and how these strategies can be imparted to poor learners through intervention. A wide range of metacognitive strategies have been identified in LLS studies as important LLS associated with language learning achievements (e.g., Gu et al., 2005; O’Malley & Chamot, 1990; O’Malley et al., 1985a, 1985b; Victori, 1999). These strategies include, but are not limited to, planning, advanced organizers, directed attention, self-management, delayed production, self-monitoring, self-evaluation, to name only a few. Metacognitive strategies are higher-order executive processes, without which students are essentially learners without direction and ability to review their progress, accomplishments, and future learning directions (Brown, Bransford, Ferrara, & Campione, 1983). Successful learning involves appropriate deployment of not only metacognitive strategies, but also cognitive strategies, among other types of strategies. Cognitive strategies work directly on incoming information and manage it to enhance learning. In the LLS literature, cognitive strategies identified consist of resourcing, repetition, grouping, deduction (applying rules to understand language), imagery, keyword method, inferencing (using information to predict outcomes or complete missing parts), summarizing, translation, contextualization, and so on (e.g., O’Malley & Chamot, 1990; O’Malley et al., 1985a, 1985b). These cognitive strategies bear striking similarity with the mental processes, such as organization, imagery, inferencing, elaboration, and deduction in Anderson’s theory (1983). The importance of organization has also been addressed in empirical LLS research (e.g., Lee, 2002; Raimes, 1985; Sasaki, 2000, Victori, 1999; Zamel, 1983). Organization plays an important role in connecting related ideas for learners to elaborate their schemata (elaboration). Anderson (1985) views organization or grouping as an important cognitive process for comprehension. Learners segment or chunk incoming information into smaller meaningful units, such as words or phrases, for easy comprehension, depending on the difficulty level of text. As a matter of fact, such mental processes as organization, selective attention, and elaboration cannot be addressed and understood in isolation due to the inevitable interplay of these processes in human cognition. For example, through the organization process, learners chunk information into meaningful words/phrases, after which learners may selectively attend to

76

Chapter Three

(selective attention) either meaning-related aspects of words/phrases or linguistic characteristics. Imagery is one of the ways that information is stored in memory. These stored mental images allow individual learners to match patterns similar to an original figure and to identify patterns with and without supporting organization or context (O’Malley & Chamot, 1990). Imagery, as a cognitive process, may assist learners in recalling verbal materials for comprehension. Successful language learners should always be on the lookout for and expecting to find patterns in language learning. However, less successful learners may lack the pattern-seeking strategy, and need to be aware of the patterned nature of language and how to identify patterns. Imagery is a very useful strategy for vocabulary learning, as evidenced in the keyword method in foreign language learning (e.g., Weinstein & Mayer, 1986). Ikeda and Takeuchi (2006) also noted the use of imagery by effective readers in their study. Another important cognitive process in Anderson’s theory is elaboration, which refers to relating incoming information to concepts or ideas already in memory. Anderson (1983) pointed out that elaborated memory structures are crucial for recalling and activating memory structures. The following important influences of elaboration have been discussed in O’Malley and Chamot (1990, p. 51). They may occur by: 1) redirecting activation away from interfering paths and toward paths which lead to the target concept; 2) spreading activation toward concepts that were part of the study context; and 3) enabling a reconstruction of the original text through inferences based on information available at the time of recall. Through these influences of elaboration, learners’ memory for concepts can be enhanced if the number of related ideas present is increased either at the time of study or at the time of recall. As discussed above, such concepts in Anderson’s ACT as declarative knowledge, procedural knowledge, and the three stages for proceduralization have lent support to describing LLS within the context of the cognitive theory. However, proceduralization does not dismiss the fact that both declarative knowledge and procedural knowledge intermingle in learners’ strategy use (Pressley & Harris, 2006). Learners may be able to use their LLS automatically to the extent they are less or not conscious of the existence of LLS. This means their LLS are highly proceduralized (Grenfell & Harris, 1999) and may have become learning skills (Afflerbach et al., 2008). But before they are proceduralized and

Theoretical Framework for the Present Study

77

reach the third stage (autonomous stage), they have undergone the other two previous stages, namely the declarative stage and the associative stage. There seems to be a consensus in the LLS literature (see Chamot, 2005) that learners may be conscious of their LLS at the early stage of learning and eventually use them unconsciously. However, when learners are made aware of their strategy use through training, LLS may be “declared” and exist in memory in an explicit manner (Grenfell & Harris, 1999). Similarly, Chamot (2005) also noted that most learners can call their strategies to conscious awareness. For example, think-aloud has been used as a method for learners to recall their strategy use in LLS research (e.g., Chamot & ElDinary, 1999; Gu et al., 2004, 2005; Macaro, 2001; Raimes, 1985; Zamel, 1983). With proper training, learners are generally able to describe their strategy use. In summary, LLS research has relied for the most part on Anderson’s theory, with which LLS researchers have developed a better theoretical understanding of how individual learners’ memory works and how new information is stored and retrieved. Based on detailed descriptions of discrete strategies (e.g., metacognitive and cognitive), researchers hope to intervene in learners’ use of these strategies to achieve better academic results and development of literacy as a whole. In section 3.2.2, I will discuss how mental processes can be modified based on Feuerstein’s theory of structural cognitive modifiability (SCM) and mediated learning experience (MLE). The above-mentioned cognitive approaches (e.g., Anderson, 1982, 1983) can provide us with theoretical explanations of how language is learned cognitively and taught in relation to LLS. First, learning a language is regarded as cognitive skill learning. Learning a language resembles the learning of other skills, e.g., playing badminton and driving. Second, LLS, or the behaviors and thoughts of a learner to influence his/her information processing, can be seen as complex cognitive processes (e.g., O’Malley & Chamot, 1990; Weinstein & Mayer, 1986). Third, cognitive processes such as LLS are modifiable through interventions, as theorized by SCM (Feuerstein, 1990). It is important to note that though the present study views LLS as intentional metacognitive and cognitive processes that enhance learning, it does not mean that only cognitive processes are involved in the process of language learning. Viewing LLS as cognitive processes may offer some merits, one of which is a detailed process view of learning, as discussed by O’Malley and

78

Chapter Three

Chamot (1990). However, certain drawbacks must be pointed out. The cognitive process view on learning only focuses on individual cognitive processes and does not allow for the influences of contextual factors in learning. During the process of learning, tasks, situations, and other factors all play a role. The cognitive approach focuses exclusively on processes such as planning, elaborating, inferencing, at the cost of other factors. The cognitive conceptions prevalent in the LLS research only lead one to assume that language tasks and contexts are generalizable (Donato & McCormick, 1994). Viewing LLS only as cognitive processes neglects the fact that learning in general and language learning in particular take place in real tasks and contexts. They constitute situated activities that are continually developing and influencing learners’ employment and learning of strategies in the classroom. Second, the cognitive view on learning overlooks the importance of socio-cultural aspects of learning (Norton & Toohey, 2001). According to Vygotsky (1978, p. 57), “every function in the child’s cultural development environment appears twice: first, on a social level, and later on the individual level, first between people, and then inside the child.” Social-cultural views on learning maintain that social interactions and cultural institutions play important roles in individual learners’ cognitive growth and development. It is the sociocultural views that allow us to look at LLS in real contexts where learners employ them. In this regard, Gao (2007) has also noted the recent theoretical development of LLS grounded in socio-cultural theories. In the next section, I will discuss the social-cultural views on learning.

3.2 Socio-cultural views on learning In section 3.1, cognitive perspectives on learning were examined in detail in relation to LLS. From those perspectives, individual learners are perceived to employ certain metacognitive and cognitive processes and strategies when learning (e.g., reading & writing). Therefore, a cognitive view on learning tends to highlight the commonalities in terms of the strategies learners use. However, limiting our understanding of learning strategies only to a cognitive view is to overlook the contributions that a socio-cultural perspective on literacy can make. As discussed previously, cognitive views bring about the problem of a decontextualized view on learning and tend to treat the use of strategies by learners invariantly, ignoring where, how, why, and when the learners employ the strategies. Literacy cannot be acquired simply as a cognitive process; it is a socially rooted process. Every member is influenced in multiple ways by the society in which reading, writing, and language interrelate with the

Theoretical Framework for the Present Study

79

workings of power and desire in social life (Gee, 1990). Therefore, the socio-cultural views on literacy can expand our understanding from decontextualized texts, skills, and individuals to situated literacy practices, performances, and contexts (Kucer, 2005, 2009). Writing acts, as literacy practices, are repeated and patterned literacy events within a social group. Writing presents a social activity that is perceived as a source of engagement for young writers to be connected to their experiences in the classroom community (Boscolo, 2008). Luke (1995, 1998) maintains that there are no private acts of literacy, only social ones. In this regard, writing is not simply individual learners’ acts, but also social acts impacted on by the members of the group and required in the social group. Accordingly, writing strategies, as LLS, are not only cognitive processes that help enhance writing, but also acts socially influenced in multiple ways that the group membership requires. In this light, the classroom, the locus of learning, is a physical and social venue where learners not only learn a language but also learn how to learn through meaningful social mediations in order to become effective future learners or members of the group (e.g., Coyle, 2007; Walters, 2007; for a detailed review, see Grenfell, 2007). Social mediations, which will be elaborated in 3.2.1, play a significant role in constructing meaningful interactions among learners and in generating higher mental processes (Vygotsky, 1986). Similarly, Gee (1996) also maintains that reading and writing skills (as literacy practices) themselves are not enough for children to become literate. Literacy practices are always interwoven into larger social practices, or what he calls Discourses. He defines Discourses as “ways of behaving, interacting, valuing, thinking, believing, speaking, and often reading and writing that are accepted as instantiations of particular roles by specific groups of people” (Gee 1996, p. viii). All literacy practices are tied to particular Discourses, which in Vygotsky’s view can be understood as social mediations. In a similar vein, Kern (2000) asserts that the process of becoming literate means not only apprenticeship with texts but also apprenticeship in particular ways of being. People do not only engage themselves in abstract processes when reading and writing, but their readings and writings are also influenced by their values held, other members in the same group, and the like. Thus, reading and writing entail more than the acts of reading and writing per se; rather, they are social practices. In a nutshell, the aforementioned advocates of social-cultural perspectives have made it clear that literacy bears a strong social orientation.

80

Chapter Three

3.2.1 Vygotsky’s social mediations and zone of proximal development (ZPD) Socio-cultural views on literacy maintain that every learner is an active member in the social mediation process. This view provides us with a framework for understanding how learners become competent members of a language learning community through social interactions (mediations) and in cultural institutions, such as schools and classrooms. Young learners spend most of their study time in the classroom. The classroom can be perceived as a main social arena where their beliefs, values, strategies are formed and mediated upon by the social practice of the members of the community, such as their friends, teachers, and so on. Therefore, central to the issue of development and use of strategies are the mediations between learners and the teacher and among learners in the classroom. Strategy-based instruction is such a phenomenon in point. Learners develop their problem-solving skills through social mediations with their co-learners and teachers. Specifically, Vygotsky’s (1978, 1986) concept of Zone of Proximal Development (ZPD) plays a central role in theorizing how social interactions can contribute to learning in general and development of LLS in particular. In this section, I will examine how LLS are developed out of social mediations in communities of practice among learners and teachers. Put another way, strategies are developed in the classroom where individual learners are initially inexperienced and unaware of LLS, but through mediations with other members gain more experience and independence in the use of LLS. In recent years, LLS researchers have begun to examine how language learners’ strategy development and use evolve from a sociocultural perspective. For example, peer review can be seen as a socially constructed activity within learners’ respective ZPD (e.g., Guerrero & Villamil, 2000; Hu, 2005; Min, 2006), with a socio-cultural orientation, as reviewed in chapter 2. Vygotsky (1986) believes that a typical learning activity is a socially meaningful cooperative activity for learners, i.e., social mediations. The acquisition of new cognitive functions and learning abilities stem from this interpersonal interaction, after which they are internalized and transformed, becoming learners’ cognitive processes. With time, experience and continued meaningful assistance, learners internalize the strategies and can deploy them independently. In other words, their LLS are developed through mediations. In terms of the development of each individual,

Theoretical Framework for the Present Study

81

mediation serves multiple purposes. First, it works as a connecting mechanism between the social level and the cognitive level. Second, mediation serves to link between individuals’ behaviors and the external phenomena at the social level. Third, an individual’s cognitive processes and strategies are modified through mediations. In other words, learners progress in their strategy learning from an early stage where they are mediated upon through meaningful activities or situations to a later stage where they are more independent. To illustrate mediations in relation to writing strategies, Guerrero and Villamil (2000) found that their student writers eventually assumed responsibility and adopted a more active role in their strategy use and writing after two mediation sessions with their readers. The above discussion has demonstrated the importance of mediations to account for LLS from a socio-cultural perspective. According to Vygotsky (1981), there are three major classes of mediators: material tools, psychological tools, and other human beings. Both the material tools and other individuals only have indirect influences on cognitive processes as they are directed at the processes in nature. In Vygotsky’s view, the psychological tools are the symbolic artifacts that mediate social interactions among human beings and the cognitive processes or strategies within the individuals. These tools include, but are not limited to, a wide range of symbolic artifacts, such as languages, symbols, signs, formulas, and graphic devices (Wertsch, 1981). These tools, through the use of which individuals’ cognitive processes and strategies can be impacted, controlled, restructured, or improved by virtue of learners’ interactions with each other, can be viewed as social ones. This kind of interactions constitutes mediations. To Vygotsky, it is the psychological tools, i.e., symbolic artifacts (e.g., languages, symbols) that serve as the primary mediators, connecting the social level and the cognitive level. However, for Feuerstein (1990), whose theory will be discussed in 3.2.2, human beings are the primary mediators. As noted by Kozulin (1994), “the role of the human mediator is not fully elaborated within Vygotsky’s theoretical framework” (p. 284). Unlike Vygotsky, Feuerstein and his associates have emphasized the presence and major role of human mediators in learning. In Feuerstein’s theory, the mediators are the human beings who intervene between the individuals or groups and the environment with the purpose of creating conditions (see 3.2.2) for the individuals’ cognitive growth. Similar to Feuerstein, Kozulin (1994) defined human mediators as those human beings that create cognitive prerequisites for making direct learning more effective. The mediators can be peers, teachers, parents or any other human beings.

82

Chapter Three

Despite their differing definitions of mediator, Vygotsky (1981) and Feuerstein (1990) share some commonalities in their views. Both theorists stress the important role that social interactions (social level) play as the fundamental influence on cognitive processes and strategies (cognitive level). The second is seen in the use of the terms tools (Vygotsky) and instruments (Feuerstein) with reference to mediation. Though used and defined somewhat differently, both terms refer to the mechanism that brings about change to cognitive processes and strategies. Both Vygotsky’s and Feuerstein’s theories on mediation are useful in explaining how social interactions (mediations) help individuals learn from a socio-cultural perspective. Though the underpinnings of their definitions and theories differ, they are both useful for my following discussion on the notion of ZPD, which is a hypothetical area within a continuum of increasing participation, relative to a learner’s personal limitations (Hickey & Granade, 2004). There is a distance between what learners can achieve with the assistance of others and can accomplish without it. The zone ends when the assistance of the others no longer results in any increased success. During the initial stage of learning, more competent or experienced learners or teachers, who in this context can be perceived as mediators, are directly involved. They structure the literacy or learning event such that young or less experienced learners can participate in a meaningful way. This structuring is called scaffolding. For example, during my first few years of schooling, my mother would show me how to hold a pen properly. I was always reminded of writing words carefully and clearly so that others could understand what I wrote. She even detailed the strokes of characters when she saw the skewed characters I wrote. At school, my teacher would demonstrate to my classmates and me how to write a proper functional text, such as an invitation letter. She discussed with us why we wanted to have an invitation letter and what it should be like, in terms of language and structure. These examples show how the more experienced person helped less experienced ones improve in learning both at home and in the classroom. When the young learner begins to learn and internalize content knowledge (e.g., particular aspects of language) and LLS, the mediator helps modify or begin to remove the support structure through assessing the learner’s current ability. The purpose of gradual decrease of support is to ensure that the inexperienced learner becomes more independent with the help and

Theoretical Framework for the Present Study

83

support of another experienced individual, i.e., the mediator. Whenever the experienced individual perceives that the young learner needs support in new aspects of learning, s/he provides support through scaffolding and mediations. This pattern of interaction ensures that the young learner participate in real learning activities as fully as possible. Coyle (2007) showed how the teacher’s scaffolding and mediation influenced the students’ strategy use both implicitly and explicitly. The analysis demonstrated how development of LLS was facilitated with the teacher’s scaffolding in a strategic classroom. The students also learned to scaffold and mediate each other in their development of self-regulation with appropriate strategy use in the process. With regards to the teaching of writing in the classroom, for example, the teacher may initially assume most of the responsibility. In other words, the teacher may detail all the words, grammar, sentence structures that learners may need to use. S/he may also need to think aloud for the learners in order to show how the strategies (e.g., goal setting, planning, and peer revision) work. The teacher will spell out why certain text types need certain structures to achieve certain writing goals. Gradually, the teacher will release more responsibility to the learners so that they eventually reach their potential and gain full independence in learning in general and writing in particular. Donato and McCormick (1994) examined how mediation helped develop students’ LLS from a socio-cultural perspective. They maintained that the development of LLS is mainly a by-product of mediation and socialization into a community of language learning practice. In their case study, they looked at how the use of a portfolio assessment procedure mediated upon the students’ growth and development in learning strategies. In this study, the students were asked to document and reflect upon their own learning growth and development through a portfolio assessment project. The instructor provided written comments, encouraging strategy use where it was found in the portfolio. As explained by the researchers, the portfolio assessment represented an attempt to create a community of language learning where the students had ample opportunity to interact with their instructor and among themselves. Over time, their strategic approaches to language learning would be improved. The analysis showed that the students’ learning became more goal-driven and their goals more specific through the project. It was also found that the students became more aware of their own strategy use. For example, conversing with others was not initially considered as a strategy. However, the students began to realize

84

Chapter Three

that conversing with others had become an often-used strategy through their self-assessment. The analysis also indicated that the students became more critical of their own strategy use; they documented the benefits of LLS in their end-of-semester written reflections.

3.2.2 Feuerstein’s structural cognitive modifiability (SCM) and mediated learning experience (MLE) The social-cultural view not only helps deepen our understanding of what LLS are, how they are learned and used, but also provides a theoretical framework for conducting interventions in the classroom successfully. In this section, I will discuss why learners’ strategic repertoires are modifiable according to Feuerstein’s (1990) theory of structural cognitive modifiability (SCM). Though SCM may be more cognitively oriented, I will discuss it in this section. This is for the sake of continuity as his theory of mediated learning experience (e.g., Feuerstein, 1990; Feuerstein, Klein, & Tannenbaum, 1999) will also be discussed in this section. The modifiability of cognition is at the heart of Feuerstein’s theory of SCM. Feuerstein (1990) argues that a learner’s capacity to learn not only depends on his/her genetic endowment but also the modifiability of the cognitive functions of human minds. “All human beings need to be considered as open systems, liable to be meaningfully modified by environmental intervention” (Feuerstein & Rand, 1997, p. 5). According to the SCM theory, even children with cognitive deficiencies have every chance for cognitive modification and positive development when the conditions are favorable (Falik, 2000). It is important to note that SCM emphasizes the modifiability of human cognition, which is comprised of perception, memory, learning, and thinking by an individual. Numerous empirical studies have shown evidence that human cognition is indeed modifiable (e.g., Tzuriel, 1996, 1998; Tzuriel & Hatzir, 1999; Tzuriel & Kaufman, 1999, cited in Tzuriel, 2000). The theory is not only applicable to children or individuals with disabilities in learning but also to those normal and high functioning ones. In Tan’s (2000, cited in Seng, Pou, & Tan) study, a polytechnic student in Singapore was found to exhibit poor planning behavior in organizing studies and life. Upon intervention, he not only exhibited planning behaviors when solving problems in class but also organized, planned and managed his time and life more effectively. It is worth noting that the cognitive processes (e.g., planning, selective attention, and monitoring)

Theoretical Framework for the Present Study

85

discussed in 3.1 overlap with the cognitive functions suggested by Feuerstein (1990). The theory of SCM implies that cognitive processes, strategies and skills are modifiable through specific and focused intervention, which Feuerstein has termed as mediated learning experience as discussed below. According to Feuerstein, there are two ways that that human cognition can be modified. The first is exposure to stimuli that affect the individual in a randomized, direct, and non-mediated way (Feuerstein & Feuerstein, 1999). The second is called mediated learning experience (MLE), which refers to the “quality of the human-environment interaction that results from changes introduced in this interaction by a human mediator who interposes him/herself between the receiving organism and the sources of stimuli” (Feuerstein, 2003, p. 23). In this view, the mediator’s role is to help the student to learn, rather than “teaching” knowledge in a traditional sense. The mediator can be persons such as parents, teachers, peers and significant others who create the necessary conditions for intervention to occur. In other words, MLE stresses the importance of the human mediator’s presence as discussed in 3.2.1. MLE, incorporating both cognitive and socio-cultural perspectives, can best explain how successful strategy training can be conducted theoretically and in practice. While MLE emphasizes the importance of interactions between mediator and mediatee, not all human interactions are mediational (Feuerstein et al., 2003). Three important parameters are generally seen as essential conditions for mediation to occur: intentionality, mediation of meaning, and mediation of transcendence (Feuerstein & Feuerstein, 1999). Intentionality, which refers to the mediator’s intention to focus on the leaner’s needs and ability to shape the task according to his/her needs, constitutes the first necessary condition for successful MLE. The intentionality of the mediator is different from that of a teacher. Traditionally, a teacher’s duty in the classroom is seen as the teaching of knowledge (i.e., content) to learners. However, from the MLE point of view, the mediator should not only be concerned with solving the problem with the learner together but, more importantly, demonstrate how the learner should approach solving the problem. This condition suits the needs of SBI well. To be specific, the teacher in the classroom should be more focused on teaching what the strategies are to solve the language

86

Chapter Three

problems faced by the learners, and how they can apply the strategies to the task at hand with the resources available to enhance their learning. .

The next necessary condition for successful MLE is mediation of meaning, which requires the mediator (the teacher or other learners) to help the learner to see what has been achieved and how the achievements have been made. In the context of SBI, mediation of meaning is meant for the mediator to lead the learner to reflect upon his/her own learning strategies that have been employed to solve a problem successfully. For example, after the learner has performed a writing task using the strategies introduced by the teacher, the teacher should help the learner realize what strategies s/he has employed for completing the task and how, where and when the strategies have been used. The third condition, transcendence, is a necessary precursor to learners’ independence in the use and learning of LLS. Human beings can transfer lessons learned from one experience to another. In other words, transcendence means “helping the learner make associations to related past experience and project himself or herself into the future” (Guerrero & Villamil, 2000, p.53). This condition means that the leaner needs to bridge the experience and strategies learned in the current task with new tasks. There are at least two purposes of LLS training with reference to transcendence. First, it can mean to apply strategies learned to solve problems in one context to problem solving in other contexts. Second, it can mean “learning to learn”, that is, acquire and develop more strategies by learners themselves. With these conditions, MLE describes that a special quality of interaction between a mediator and a mediatee, and the quality of interaction between the individual (mediatee) and the environment via an intentional human being (mediator) plays a pivotal role in the cognitive development of the individual (Feuerstein, 1990; Feuerstein et al., 1999). MLE is a theory that has been used as a theory for conducting interventions (Tzuriel, 2000), through which modification of individual cognitive processes can occur. The three conditions will be integrated into the design of SBI for the present study.

Theoretical Framework for the Present Study

87

3.3 Multi-dimensional views on language learning and literacy In the previous two sections (3.1 and 3.2), the cognitive and socio-cultural perspectives of LLS were examined in detail in order to situate the present study on a broader theoretical foundation. LLS are not only a set of cognitive processes but also are mediated upon socio-culturally. The cognitive and socio-cultural approaches should not be perceived as incompatible with or inimical to each other. Rather, both cognitive and socio-cultural processes are interacting with each other continuously during learners’ literacy development. In fact, LLS researchers (e.g., Oxford & Schramm, 2008) have argued for the importance of the interactions between both these two over-arching theories in order to gain a better understanding of learner’s strategy use. In other words, the cognitive and socio-cultural perspectives should not be treated as two parallel SLA worlds or two conflicting (incompatible) perspectives (e.g., for detailed reviews, see Mason, 2007; Zito, Adkins, Gavins, Harris, & Graham, 2007; Zuengler & Miller, 2006). Instead, the two perspectives should be taken together to understand language learning in a larger theoretical framework. In chapter 1, it was also made clear that the present study examines LLS in relation to learners’ literacy development. Therefore, a multi-dimensional view on literacy is central to explaining how learners develop and use LLS. The main theoretical framework presented here is Kucer’s (2001, 2005, 2009) dimensions of literacy. Drawing on literacy research from different disciplines (e.g., linguistics, cultural studies, and psychology), Kucer (2001, 2005, 2009) and Kucer and Silva (2006) maintain that literacy should be conceived as dynamic and multi-dimensional in nature. Kucer notes that each discipline tends to define and create literacy in its own image. For example, cognitive psychologists are interested in the mental processes of the learner for meaning making, whereas researchers following socio-cultural approaches view literacy as expression of selves as members of socio-cultural groups. Therefore, it is highly important to unify literacy research in light of the development of the multiple disciplines. He argues that to become literate, learners must know how to “effectively, efficiently, and simultaneously control the linguistic, cognitive, socio-cultural, and developmental dimensions of written language in a transactive fashion” (Kucer, 2009, p. 5).

88

Chapter Three

As illustrated in Figure 3.2, four dimensions – cognitive, linguistic, sociocultural, and developmental – constitute each literacy act. Kucer (2001, 2005, 2009) metaphorically describes the role of each dimension as code breaker (linguistic dimension), meaning maker (cognitive dimension), text user and critic (socio-cultural dimension), and scientist and construction worker (developmental dimension). At the centre of the literacy act is the cognitive dimension, which describes the cognitive skills or processes that learners go through.

developmental sociocultural linguistic cognitive literacy events

Figure 3.2. Dimensions of Literacy (Kucer, 2009, p. 5)

Kucer’s cognitive dimension bears striking similarity with the cognitive views on LLS as discussed in 3.1. In this dimension, meaning is constructed by the mental processes, strategies, or procedures the learner engages in. Kucer (2005, 2009) explains that, for example, perception is traditionally conceived as a one-way process in which information recorded by the eye is sent to the brain and processed there. However, perception should be understood in more dynamic and constructive ways, based on perception experiments. Smith (1994, 2004) suggests that perception involves a two-way process. The brain does not only receive

Theoretical Framework for the Present Study

89

information passively, but it can determine and require the eye to select information that is unfamiliar, important or relevant in an active manner. Learners are found to deploy LLS such as selective attention (Carrier, 2003; Ikeda & Takeuchi, 2003) and monitoring (Gu et al., 2005; Nisbet & Shucksmith, 1986; Rubin, 1981; Vandergrift, 2003a). They actively process the information perceived through their eyes. Both reading and writing strategies are described as cognitive processes in this dimension, which are similar to those covered previously in section 3.1. To Kucer, strategies are the cognitive processes or behaviors that the learner engages in to create meaning through the written discourse. For example, readers and writers constantly predict meanings when they are reading or writing. They also monitor and evaluate their understanding in the process. Kucer (2009) also notes that the orchestration of strategies occurs during learning. For example, he explains “more than one strategy may occur at any given moment and these strategies need not operate in a particular sequence” (p. 120). The linguistic dimension in Kucer’s multi-dimensional view is concerned with the systems of language. “Language is the vehicle or avenue through which ideas are constructed during reading and writing” (Kucer, 2009, p. 17). Pragmatics, text type, semantics, syntax, morphology, orthography, and graphophonemics all play a role during learning. Various systems of language within a text form a web of meaning. Learners need to know about the language systems to “crack the code” when they are dealing with text. In relation to LLS, language is a primary means for learners to learn or deploy LLS to make their learning more effective. For example, learners need to use language when they learn to describe what LLS are, how they are used, and why they are chosen. Their linguistic competence promotes or impedes their development and use of LLS. As was pointed out in 3.2.1, language, as a kind of important psychological tool (Vygotsky, 1978, 1986; Wertsch, 1981), serves as a primary mediation tool, through the use of which modifications and improvements of individual learners’ cognitive processes (cognitive dimension) can take place. Some learners may use their LLS more appropriately and successfully if their language competence can well support their strategy use. On the other hand, some may not be able to use the same strategies as successfully as their successful counterparts due to their poor linguistic competence (i.e., linguistic dimension).

90

Chapter Three

The third dimension is the socio-cultural dimension. As noted by Kucer (2005, 2009), literacy events are more than the cognitive processes that individual learners go through for meaning making. This dimension of literacy maintains that whereas cognitive views on learning strategies are used to explain individual differences in learning, socio-cultural theories (Vygotsky, 1962, 1978) can best describe how individuals learn through their social interactions with each other. In a typical learning situation, learners acquire knowledge through social interactions with others. This dimension is similar to the socio-cultural views on learning as discussed in 3.2. Literacy learning is also a social act. It is through the social interactions (e.g., mediations) among learners that their LLS are developed in this dimension. Many empirical LLS studies have addressed the importance of this dimension (e.g., Coyle, 2007; Donato & McCormick, 1994; Guerrero & Villamil, 2000; Hu, 2005). Likewise, Lantolf (2000, 2006) and Lantolf and Thorne (2006) also noted the importance of the union of socio-cultural theories and cognitive theories in L2 research. In other words, individual learners learn how to use LLS (i.e., cognitive dimension) through the use of language (i.e., linguistic dimension) through mediations (i.e., socio-cultural dimension) with other learners. Therefore, SBI can be seen as a process where the three dimensions are well connected with each other. Lastly, encompassing the cognitive, linguistic, and socio-cultural dimensions is the developmental dimension. This dimension describes the relationship among the other three dimensions. The development of literacy is a neverending process. “Individuals may encounter literacy events that involve using literacy in new and novel ways” (Kucer, 2009, p. 6). With the widening range of learning contexts and experience, the three dimensions are continuously extended and developed. On the one hand, the learner gains more control over them during this process. On the other, the leaner needs to develop more to cope with the increasing demand due to the widening range of learning contexts and experiences. Four principles are involved in this dimension. First, the learner is perceived as a scientist and construction worker. It reflects the learner’s role as actively constructing linguistic, cognitive, and socio-cultural competence through the use of LLS, mediations, and so on. Literacy learning is a process in which the three dimensions (linguistic, cognitive, and socio-cultural) continually transact with each other in a complex manner, while moving forward together. LLS and mediations work as the tools, through the active use of which the development of the three dimensions can occur. Second, this dimension is a recursive process. The literacy learning process is both a

Theoretical Framework for the Present Study

91

progressive and regressive process by which the learner may advance or move back. Along the process, the learning of LLS may also move forward or move back. As was discussed in 3.1.2, there exist three stages for the development of LLS: the declarative stage, the associative stage, and the autonomous stage. Though the process is generally forwardmoving, there may be problems in the process and the development stops or moves backward. For example, the learner may have developed a cluster of strategies for dealing with a specific language learning problem. There may be ineffective rules or ill-productions (Anderson, 1983) that impede the process of optimization of the production rules. The illproductions may not be removed easily. As a result, the proceduralizations may stop due to the ill-productions, or the learner’s learning may move backward as the ill-productions prevent the learner from resolving his/her problem(s) effectively. Third, the adult or a more experienced individual works as the demonstrator, mediator, and guide. Initially, the adult or the more capable person needs to demonstrate to the learner how learning takes place cognitively, linguistically, and socio-culturally. As has been made clear in 3.2.1, young children’s literacy learning is dependent on the scaffolding created by the adult or the more capable person, who mediates on the young learner. This principle addresses the need for a mediator in the learner’s literacy development process as well as the learning of LLS. Fourth, negotiating meaning is involved (Kucer, 2005, 2009). All forms of communication are needed in order for the learner to learn. The young learner may still not be competent in the three dimensions yet. Therefore, the mediations between the adult and the learner must be full of cues or signs, e.g., body language, pictures, and symbols. In Vygotsky’s (1981) view, these forms are the material tools which impact on individuals’ cognitive processes. However, to extend his view more broadly and in line with Kucer’s (2005, 2009) multi-dimensional view, the material tools also impact on the linguistic and socio-cultural dimensions in addition to the cognitive dimension. To sum up, the developmental dimension denotes that language learners’ mastery of LLS is continuously evolving in all the four dimensions. The whole process of LLS development is a neverending process. Kucer’s framework provides a more comprehensive view on literacy in relation to LLS. However, this framework is not without problems. According to Kucer’s theoretical discussion of and positioning of the different dimensions, as shown in his figure, it can be inferred that it is the

92

Chapter Three

cognitive dimension that is firstly developed or worked on among all the other dimensions, followed by the linguistic, and socio-cultural. Kucer (2009, p. 5) notes that “at the centre of the literacy act is the cognitive dimension, the desire of the language user to explore, discover, construct, and share meaning.” It can be argued that in a literacy act, the development of the three dimensions may occur simultaneously and interactively. It is difficult to tell which dimension will lead in the process as the configuration of the dimensions may vary at different stage of literacy learning as well as in different mediations. So sequencing the different dimensions of literacy may lead us to a problematic understanding of how the different dimensions of literacy interact. A model of multi-dimensional framework of literacy similar to Kucer’s (2005, 2009) was also proposed by Kern (2000). There are three dimensions: the linguistic, the cognitive, and the socio-cultural dimensions. In fact, the three dimensions in Kern (2000) resemble those (the linguistic, the cognitive, and the socio-cultural) dimensions in Kucer (2005, 2009). However, the two frameworks are different in several ways. First, Kern’s three dimensions are not only overlapping and interdependent, but also embedded within each other. To Kern, each perspective only provides a partial view of literacy. The three perspectives must be taken into consideration together to illuminate literacy’s multiple aspects. Figure 3.3 demonstrates the linguistic, cognitive, and socio-cultural dimensions of literacy and their interrelationships as conceived by Kern (2000, p. 38). In contrast, Kucer’s (2005, 2009) dimensions, though recursive, do not interact with each other in a dynamic way as Kern’s dimensions. For example, Kucer’s framework indicates that each dimension may develop in its own way leaving out the interactions among the dimensions. The dimensions bear no influence on each other. Kern’s framework suggests that all the three dimensions interact and depend on each during the learner’s literacy learning process. Each dimension only offers a partial picture of literacy. Therefore, Kern’s framework may explain better the relationships among the different dimensions as compared to Kucer’s. Second, Kucer’s (2005, 2009) framework includes a fourth dimension, developmental dimension, whereas Kern’s framework does not. In this regard, Kucer’s model may have demonstrated better how literacy learning evolves and progresses. It is the developmental dimension that binds all the other dimensions together and promotes the development of all the dimensions in the literacy learning process. In this regard, it seems that

Theoretical Framework for the Present Study

93

Kern’s framework only details clearly the intricate relationships of the multiple dimensions, but does not describe how progress is made.

Sociocultural x Collective determination of language uses and literacy practices x Interweaving of literacy practices with other social practices x Apprenticeship into ways of being (social acculturation, acquiring Discourses, joining the literacy club) x Social and political consciousness: problematizing textual and social realities x Awareness of dynamism of culture and of one’s own cultural constructedness

Linguistic x Lexical morphological, syntactic, semantic, pragmatic knowledge x Familiarity with writing system and graphic and organizational conventions x Awareness of interdependencies at all levels (orthography, lexicon, sentence, paragraph, text) x Awareness of relationships between oral and written language (including awareness of distinction between medium and mode of expression) x Familiarity with genres and styles

Cognitive/Metacognitive x Existing knowledge (schemata) – allowing a person to establish relationships among pieces of information and to predict, infer, and synthesize meaning o Declarative knowledge – the ‘what’ – facts, ideas, stories embedded in cultural contexts o Procedural knowledge – the ‘how’ – strategies for reading, writing, and understanding, also embedded in cultural contexts x Ability to formulate and discern goals and purposes – including planning, monitoring, and revising – in line with cultural norms x Ability to create and transform knowledge

Figure 3.3. Summary of dimensions of literacy (Kern, 2000, p. 38)

Having examined both Kucer’s and Kern’s model of literacy, it is important to note that a great deal of previous and current LLS research (e.g., Carrier, 2003; Lee, 2002; Hong-Nam & Leavell, 2006; Ikeda & Takeuchi, 2006; Magogwe & Oliver, 2007; O’Malley & Chamot, 1990; O’Malley et al., 1985b; Raimes, 1985; Zamel, 1983) reviewed in chapter 2 has tended to focus exclusively on the cognitive dimension of literacy. In this view, LLS are narrowly equated with cognitive processes that learners must use invariantly. The teaching of LLS is framed over-simplistically in terms of the cognitive development of learners. The contributions of the linguistic, socio-cultural, and developmental dimensions are largely neglected. Teaching and learning in the multiple dimensions of literacy is

94

Chapter Three

essential for learners to become successful in a linguistically and socioculturally diverse world (Anstey & Bull, 2006). A fruitful study of LLS requires a reconceptualization of the dimensions on which literacy and LLS work. If LLS are examined from a multi-dimensional view, each dimension should be addressed in LLS research as each dimension plays an important and yet different role in literacy learning in general and acquisition and use of LLS in particular.

3.4 Conclusion In this chapter, I first examined the cognitive theories with respect to LLS. It was shown that much of the existing research was conducted based on the assumption that LLS possessed a strong cognitive origin. In particular, Anderson’s ACT can effectively analyze how individual learners deploy their LLS cognitively. However, cognitive theories suffer from the problem of a decontextualized view on learning. Therefore, socio-cultural theories of learning, in particular, the notions of ZPD and MLE, were also examined to provide a better understanding of how LLS are taught, learned, and improved in the classroom. It was also noted that different approaches have their own strengths and weaknesses. Each approach only offers a partial view on learning and literacy. Then, Kucer’s (2005, 2009) multi-dimensional framework on LLS and literacy was proposed to complement the inherent inadequacy of each approach. Different dimensions of the framework account for different aspects of literacy and LLS development while they are interacting in a complex manner.

CHAPTER FOUR RESEARCH METHODOLOGY

In the previous chapters, I reviewed the general research field of language learning strategies, and both descriptive and intervention studies of writing strategies in particular. Both descriptive and intervention studies have yielded mixed findings due to the different research methods adopted. There have been fewer intervention studies as compared to descriptive ones in classroom settings (Chamot, 2005). It was also made clear in chapter 1 that this project is focused on both describing the writing strategies used by Singapore upper primary school pupils and helping them acquire and use appropriate writing strategies effectively. In this chapter, I will give a detailed account of the research methods employed in the present study in light of the methodological issues in the LLS literature and theoretical underpinnings discussed in chapter 3. The research project includes two phases: 1) establishing patterns of strategy use, and 2) determining the effectiveness of strategy-based instruction (SBI). I will detail the research methods for each phase separately.

4.1 Establishing patterns of strategy use (Phase one) This phase was descriptive and mainly involved the administration of a writing strategy questionnaire to identify the range and type of writing strategies used by upper primary pupils in Singapore. It also aimed to map out the relationships between strategy use and language proficiency. Once the relationships are known, a strategy-based writing intervention program can be designed for upper primary school pupils. Therefore, Phase 1 will partly help inform the design of Phase 2. Phase 1 addressed the following research questions: 1. 2.

What writing strategies do upper primary school pupils use? How are the differences in strategy use related to language competence?

96

Chapter Four

4.1.1 The writing strategy questionnaire and its development Questionnaire is the most frequently used data collection method for identifying learning strategies used by learners (Chamot, 2005). Many LLS studies have used questionnaires (e.g., Chamot & El-Dinary, 1999; De Silva, 2015; Gunning & Oxford, 2014; Ikeda & Takeuchi, 2003; Lee, 2002; Lan & Oxford, 2003; Magogwe & Oliver, 2007). There are some limitations associated with questionnaires. For example, learners may fail to report what strategies they have used in the past when responding to the questionnaires. They may not understand some of the questions on a questionnaire. Therefore, care must be taken to ensure that some of the problems associated with questionnaires are resolved. There are also apparent advantages using questionnaires. It is easy to administer a questionnaire at multiple schools at different times. It is also possible to collect data from a large sample with a written questionnaire. The major instrument used for obtaining data on the pupils’ strategy use for the present study was a writing strategy questionnaire that was developed based on the previous literature (e.g., O’Malley & Chamot, 1990; O’Malley et al., 1985a, 1985b; Oxford, 1990; Petriü & Czárl, 2003). As was pointed out in chapter 2, simply adopting and employing an existing questionnaire for data collection may not be effective. The process of questionnaire development in the present study represented an effort of “grounded approach” (Macaro & Erler, 2008) in that 1) sufficient piloting was conducted; and 2) the questionnaire was developed within the context of institution type and proficiency level of the participants. The questionnaire consists of two main parts: 1) seven questions on personal data (e.g., age, gender, race, and so on), and 2) 46 questions on strategy use. All the participants’ latest English language results were also obtained from the schools. The procedures of piloting will be reported below. The writing strategies identified in a previous study with 32 pupils served as an initial pool of strategies for this phase. When selecting strategies for inclusion in the questionnaire, the research team mainly chose those that were used more frequently by the pupils. In addition, the strategies that could differentiate between successful and unsuccessful pupils were also included. As the strategies found with the 32 pupils might not be representative of the repertoire of writing strategies for Singapore primary school pupils, given the small sample size and the restrictions arising from the use of only two data collection tasks, strategies discussed in the literature were also considered.

Research Methodology

97

First, an item pool of 58 writing strategies (Version One) were shortlisted by a key member on the research team with my assistance by consulting some relevant research studies of writing strategies and some local primary school teachers. A statement was constructed to describe each strategy. The statements were sequenced following the general structure of the writing process, i.e., pre-writing, writing, and post-writing during the development period for the purpose of easy referencing among the research team members. Version Two was an outcome of refinements by all the research team members made up of experienced LLS researchers at several meetings. Special attention was paid to whether the statements under each strategy were clear and accurate enough to describe the strategy. Then, this version was reviewed by four experienced primary school teachers. Potentially misleading items and wording confusions were identified. The sequence of the items was randomized using a computer program. The randomization was to ensure that the participants would not make the same choice for the statements in the same category because they sound similar. Version Three was then developed based on these teachers’ feedback and suggestions. At this point of time, more than enough strategies were included in each category so that those that may not be loaded well statistically could be removed and there would still be enough items for the final version of the questionnaire. Version Three was then validated based on the think-aloud data collected from six pupils taking the questionnaire at a primary school. The pupils were asked to answer the questionnaire while thinking aloud in front of a member of the research team. For each statement, the participants were asked to rate how often they used the writing strategies using a Likert scale from 1 to 5, with 1 meaning “never” to 5 meaning “always”. Whenever the pupils had problems with either the language or the strategies themselves, problems were noted down. Then, improvements were further made on the questionnaire at a review meeting based on the field notes. This process resulted in a 50-item questionnaire (Version Four). The questionnaire was then piloted with 221 pupils at three grade levels (primary 4 to primary 6) in a primary school. Both item analysis and factor analysis were done to identify weak and problematic items, which were then removed. The factor structure yielded by the factor analysis did not match the structure conceptualized based on the research team’s theoretical discussion. The results were used, however, to remove a few weak items. This is in agreement with Woodrom’s (2005) observation that the results of many LLS studies using factor analysis are fraught with contradictions. Caution should be taken when interpreting the results. Therefore, the factor analysis could not serve as a main statistical method for evaluating the validity of the questionnaire. However, the factor

Chapter Four

98

analysis did provide some useful information. For example, it showed that the monitoring strategies and evaluating strategies were loaded better as one factor. Indeed, the item analysis also showed better internal consistency when they were combined. The scale alpha ranges from 0.421 to 0.781. Version Five was developed based on the statistical analyses. After removal of the less reliable items, 46 items remained in the final version. The process of the questionnaire construction from Version One to Version Five also represented a good effort in establishing content validity (for a detailed review, see Petriü & Czárl, 2003). Table 4.1 presents the scale alpha values of the item analysis of version five. The scale alpha for this version ranges from 0.526 to 0.724, which allowed me to conclude that the questionnaire was within the range of acceptability (DeVellis 1991; Dörnyei, 2007; Petriü & Czárl, 2003). However, it is important to point out that the internal consistency as measured by Cronbach’s alpha only serves as a guideline rather than the only criterion (for a detailed review, see Petriü & Czárl, 2003). More practical validation methods should also be involved, such as content validity and construct validity. In this study, content validity was considered in the process of the questionnaire development since it was an important step during the construction of the questionnaire. The procedures that were involved in the development of the first three versions of the questionnaire were to ensure the content validity. Table 4.1. Writing Strategy Questionnaire (Version Five) Item Analysis Subscale

Item #

Strategy

Scale alpha

1

I try to find out how to write good English compositions. Besides homework, I also write English compositions at home to improve my writing. I look at good English compositions in order to write well. I ask for writing tuition to improve my English writing. I ask my teacher, friends or family members to tell me what they think about my compositions and why they think that way.

.616

Self-initiation

26

3 5 21

Research Methodology

10

4 31 2

Planning

14

30

7

33

23 32

Monitoring and Evaluating

39

36 9

11

Before I write an English composition, I discuss the topic with my teacher, friends or family members. Before I write an English composition, I read about the topic. Before I start writing an English composition, I write out a plan for it. When I plan for my English composition, I think about what ideas to put down. When I plan for my English composition, I think about what words, phrases, and sentences to use. When I plan for my English composition, I keep in mind the teacher’s requirements for the composition. When I plan for my English composition, I think about how my readers will feel about my composition. When I plan for my English composition, I think about how to organize my ideas. When I plan, I keep in mind what type of composition I am going to write. Before I start writing an English composition, I plan for it in my mind, but not on paper. When I am writing an English composition, I tell myself to stick to my plan. I read my composition aloud to look for mistakes. When I am writing an English composition, I try to make my spellings and grammar right. After writing my composition, I check whether it meets my teacher’s requirements.

99

.723

.634

Chapter Four

100

22

27 24

Feedback handling

40

37 28

35

16

Revising

45 18 6 8

Generating text

29

19

20

I compare my composition with my friends’ compositions on the same topic. When I read my composition, I can tell whether it is good or bad. When I read my composition, I can tell whether my readers will like it or not. I try to remember or write down other people’s suggestions so that I can use them in the future. I carefully think about other people’s suggestions for my compositions. If my teacher, friends, or family members ask me to make changes to my composition, I ask why. I check my mistakes after I get back the composition from my teacher, and try to learn from them. When I check my English composition, I change spellings or punctuations. When I check my English composition, I change words or phrases. When I check my English composition, I change grammar. When I check my English composition, I change ideas in it. When I check my English composition, I re-organize it. When I am writing an English composition, I read back what I have written to help myself think of new ideas. When I am writing an English composition, I read other people’s writings for words, phrases, and sentences to use in my composition. When I am writing an English composition, I read other people’s writings for ideas to write down in my composition.

.526

.611

.724

Research Methodology

25

38

44

Resourcing

43

41

17

15

Socio-affective

34 42 12

13 46

When I am writing an English composition, I recall ideas that I have read in books or others’ compositions for use in my own composition. When I am writing an English composition, I recall words, phrases, and sentences from books or others’ compositions for use in my own composition. When I am writing an English composition, I re-read the teacher’s requirements to help myself think of new ideas. If I don’t know a word or phrase in English, I stop writing and look it up in a dictionary. If I have difficulty in writing an English composition, I turn to my grammar books, textbooks or writing guides for help. When I am writing a new English composition, I go back to my old compositions for help. I ask my teacher for help when I have difficulty in writing. I ask my friends for help when I have difficulty in writing. I ask my family members for help when I have difficulty in writing. I tell myself to enjoy writing.(interest/motivation -socialaffective) I tell myself not to worry when I am writing an English composition. I reward myself (for example, eating favourite food, watching TV for a while, going out with friends, etc.) for completing an English composition.

101

.548

.571

The 46 items are divided into 8 subscales/groups: self-initiation strategies (5 items), planning strategies (10 items), monitoring and evaluating

102

Chapter Four

strategies (7 items), feedback handling strategies (4 items), revising strategies (5 items), generating text strategies (6 items), resourcing strategies (3 items), and social/affective strategies (6 items). It is important to note that planning strategies have the largest number of items as compared with the other subscales, consistent with the research finding that planning consumes a high portion of the writing time throughout the whole writing process (Humes, 1983).

4.1.2 English language tests The pupils’ latest English language results were also obtained from each school. As the time of questionnaire administration was at the beginning of a new semester, there were no major English language tests in the schools. All the pupils’ previous English language semestral examination results were therefore obtained as a measure of their English language proficiency. Each pupil’s result was a composite score that was made up of the results that measured their reading, speaking, listening and writing. On the basis of these results, the relationship between their English proficiency and strategy use can be explored. It should be pointed out that the students’ English proficiency was measured by their overall English language results. As the questionnaire survey involved five different schools and three grade levels, it was not possible to administer one standardized writing or proficiency test to all of the participants. Therefore, their latest English language results (previous semestral English results) for each grade level in each school were used as a measure of their overall language competence. The research team reasoned that 1) the overall language competence of the participants reflected their writing ability to a certain extent, and 2) the semestral English results were the best available measure in lieu of their English composition result.

4.1.3 Participants The participants were from five representative government-run neighborhood primary schools in Singapore. These pupils were from three grade levels: primary 4, primary 5, and primary 6. The reason for inclusion of only these three grade levels was twofold. First, involving lower primary pupils in such a large-scale questionnaire administration could be very difficult. They may not understand certain items on the questionnaire. Second, upper primary pupils have developed a better understanding of the

Research Methodology

103

logistics of questionnaire administration and questions addressed on a questionnaire. Among the five schools, four were average neighborhood primary schools. One was a boys’ school and was rated as a better school, compared with the other four. Before data entry, I scanned through each answer sheet to spot questionable data. For example, some pupils chose the first option for every statement. Such pupils’ data were excluded because of apparent abnormality. After exclusion of unusable data, 3178 pupils’ data were used for analysis. These pupils were aged from 9 to 15 years old (M = 10.38 years). Of the pupils, 1,926 (60.60%) were boys, and 1,212 (38.14%) were girls, with forty pupils (1.26%) failing to provide their gender information on the questionnaire. As for ethnicity, 2,140 (68.02%) were Chinese, 795 (25.27%) were Malay, 143 (4.55%) were Indian, and 68 (2.16%) were others. Thirty-two (1%) pupils did not provide information on their ethnic groups. Such ethnic percentages were quite similar to those in the national census (Leow, 2001).

4.1.4 Procedures Before the administration of the questionnaire, a briefing was held for all the participating teachers in each school. Each teacher received a package consisting of the writing strategy questionnaire, guidelines for questionnaire administration, and the questionnaire administration record. Specifically, the teachers were asked to tell the pupils that there was no right or wrong answer to the questions, that they should not ask their friends as to how to respond to each statement, and that they should not guess how their teachers would like them to respond as their answers would not in any way influence their academic results in the school. They were also advised to respond based on their first reaction. In addition, the pupils were assured that their responses would not in any way influence their English language exam results. The teachers were also told to note down any problems encountered in the questionnaire administration record. The procedures were detailed in a checklist. Meanwhile, potential questions from the participants regarding the questionnaire and standard responses were also provided in the guidelines. Whenever a question or problem with respect to the questionnaire arose, it was immediately answered or resolved. The administration of the questionnaire lasted about 15 minutes for each class. A researcher or an assistant was around in each of the classes so that if there were any problems or questions, the researcher could be called upon.

104

Chapter Four

Completed questionnaires and questionnaire administration records were then collected from the teachers immediately after the administration.

4.1.5 Analyses The questionnaire data were entered into SPSS 21 through a computer program called Remark Office OMR. Descriptive statistics were obtained in order to see the overall pattern of writing strategies used by the participants. The independent variable was the participants’ English proficiency. The participants were grouped into three proficiency levels, i.e., top, middle, and low on the variable. The dependent variables were reported frequencies of strategy use. A number of One-way ANOVA analyses were performed to determine if the three English proficiency groups differed in their reported strategy use. If the pattern for the correlation between each variable and language proficiency is consistent for the five schools, it means that the particular variable is consistently correlated with language competence. In order to rule out any possible problems that may arise due to the use of the pupils’ English language results for the three grade levels in each school as the common measurement, the English language results of each grade level were first converted into z-scores and then the z-scores of each grade level were entered into the same data set as the measurement of the students’ language proficiency. Then, the z-scores were equally divided into three groups: top, middle, and bottom.

4.2 Determining the effectiveness of SBI (Phase two) Phase 2 aimed to examine whether strategy-based writing instruction would be effective in the Singapore school system. In particular, the research questions addressed in Phase 2 are: 1. In what ways does strategy-based writing instruction help pupils improve their writing competence in English? 2. In what ways does such instruction help improve pupils’ use of the target writing strategies? A quasi-experimental design was adopted in this phase. Three average neighborhood primary schools participated. It is important to point out that the present study also took into consideration the participating teachers’ involvement. As was reviewed in chapter 2, the researchers were the

Research Methodology

105

strategy instructors in most of the intervention studies. Very little attention has been paid to teacher training (Chamot, 2005; Gu, 1996; Macaro & Erler, 2008). Therefore, it is highly significant to measure the effects of intervention with the involvement of teachers as compared with those without teachers.

4.2.1 Participants All the 442 participants, aged between 10 and 11 years old (M = 10.7), were primary 5 pupils from three different government-run neighborhood primary schools (one of them was involved in Phase 1). They had not experienced any intervention with regard to writing before. Of them, 209 (47.29%) were girls, and 233 (52.71%) were boys. The racial breakdown of the pupils was as follows: 316 (71.49%) ethnic Chinese, 99 (22.40%) ethnic Malays, 16 (3.62%) ethnic Indians, and 11 (2.49%) from other ethnic backgrounds. Among all the participants, 221 pupils received SBI, and 221 attended their regular writing lessons. Given the big sample size in the study, it was difficult to have random selection and assignment of participants to form new groups for the purpose of the intervention due to a few educational constraints. First, the pupils were taught in different classes by different teachers so all the pupils were from intact classes. Second, the study spanned about one semester, which might cause logistical problems to the curriculum if randomization was done. So a quasi-experimental design was adopted as it was the best available option for the design given the situation. Six intact classes served as experimental classes, and six classes as control classes. Table 4.2 shows the numbers of participants in detail. Special attention was paid to the assignment of classes into different groups (experimental & control) so that the experimental and the control classes were as comparable as possible. For example, the classes’ English test/exam results were compared. Either the Head of the English Department or the Vice Principal of each primary school was also consulted with respect to the comparability of the classes in the school. Nonetheless, better classes often served as the control group as the participating schools would like the weaker classes to be helped through the SBI. As a result, it was often the case where weaker classes were the experimental group. Such an arrangement may put the experimental group at a disadvantage when it comes to analyzing the effects of the SBI. A one-hour lesson of writing strategies was conducted for the experimental classes for nine consecutive weeks.

Chapter Four

106

Table 4.2. SBI Writing Classes and Participants School

Experimental group

Control group

Total

School A

Class 1 42

Class 2 28

Class 3 42

Class 4 29

141

School B

40

39

38

39

156

School C

32

40

30

43

145

Total

114

107

110

111

442

The SBI lessons were conducted for the experimental pupils in the time slots reserved for writing instruction in the respective schools. The difference between the experimental classes and the control classes was the SBI. The pupils in the control group also received one-hour writing instruction that was conducted by their regular writing teachers. From each experimental class, four to five representative pupils at different English language proficiency levels as determined by their most recent English language results and with different strategy use patterns as identified by the pre-SBI questionnaire administration were chosen to serve as a case study group. They were interviewed both prior to and after the intervention in order to investigate their strategy use and change in greater details. All the interviews were audio-taped. Special attention was also paid to these pupils during class observations. There was also a case study group for each control class. As mentioned above, the pupils in the control classes were taught how to write for one hour each week during the intervention period by their teachers. Prior to the intervention, I had two observations on the writing instruction and held one meeting with some of the teachers in order to determine the teaching methods of writing used in the schools. Class observations showed that the writing instruction in these schools was mainly influenced by the genre approach and process writing. In the meeting, however, some teachers mentioned they also spent quite some time on grammar, vocabulary, and structure problems in their teaching as many weak pupils were still struggling with these problems in their writing. As for strategy-based writing instruction, they did not know what the method entailed and had never used it before. The teachers selected to participate in this research were form teachers. First, a project meeting was held for interested teachers as identified by the Head of the English Department of each school. The project team

Research Methodology

107

members explained to the teachers what entailed in such a project. The teachers were at liberty to decide whether they would like to participate on a voluntary basis. Altogether, 12 teachers were involved in the SBI, with six in the experimental group and six in the control group. Of these, 10 were female, and 2 were male. The age of the six teachers in the experimental group ranged from 27 to 55 (M = 36.7), and their teaching experience from 3 to 30 years (M = 10.5). The other six teachers in the control group ranged from 35 to 61 years old (M = 44.5), and their teaching experience from 7 to 35 years (M = 20.5). One project briefing was held for both experimental and control teachers. They were told that this research project was exploratory, and thus there would be no ground to expect the intervention to be effective. Especially, the teachers in the control group were advised that whether the SBI would be successful or not was yet to determine. They should not discuss with their colleagues in the experimental group as to what was being carried out. This was to avoid treatment diffusion during the intervention period. If the intervention turned out to be successful, strategy training and lesson plans would be provided to the teachers in the control classes. At the end of the briefing, a question and answer session was held to clear any doubts they might have. A 3-hour SBI workshop exclusive to the experimental teachers was provided immediately after the project briefing. The purpose was to equip the experimental teachers with the essential knowledge on how the SBI could be conducted successfully. The workshop consisted of three major parts: strategy rationale, teachers’ think-alouds and the SBI lesson plans. In order for the teachers to have a clear idea on how an SBI class should be conducted, the video clip of a model SBI lesson that I shot during the pilot was shown at the workshop. Then, a discussion session on the lesson was held. Finally, a Q & A session was held to respond the teachers’ inquires.

4.2.2 Materials Given the importance of the lesson plans and the many issues connected with them, this section is solely devoted to describing the design of the lesson plans. In 4.2.3, I will describe the various instruments used in the SBI.

108

Chapter Four

SBI Model for the present study As reviewed by Harris (2003), several models (see Table 4.3) for SBI can be found in the literature (e.g., Chamot, Barnhardt, El-Dinary, & Robbins, 1999; Grenfell & Harris, 1999, O’Malley & Chamot, 1990; Oxford, 1990). These various models, though differing in their details, share the same overall goals. They aim to: 1) raise learners’ awareness; 2) encourage strategy use and provide the rationale for use; 3) provide a pool of relevant strategies; 4) help learners reflect on strategy use (see Harris, 2003, for a discussion of the models). All these models represent an effort of explicit instruction. It was noted in chapter 2 that both explicit instruction and implicit instruction can be found in the LLS literature. However, explicit instruction seems to be more effective as compared with the implicit approach (Anderson, 2005; Carrier, 2003; Chamot, 2004, 2005; Duffy, 2002; Grenfell & Harris, 1999; O’Malley & Chamot, 1990). Hence, the present study adopted an explicit instructional approach. Table 4.3. A Comparison of Four Strategy Instruction Models O’Malley & Chamot (1990) 1. Students identify their current learning strategies

2. Teacher explains additional strategies

Oxford (1990) Learners do a task without any strategy training They discuss how they did it and the teacher asks them to reflect on how their strategies may have facilitated their learning Teacher demonstrates other helpful strategies, stressing the potential benefits

Chamot et al. (1999) Preparation

Grenfell & Harris (1999) Awareness raising. Learners do a task “cold” They brainstorm the strategies used. Class shares strategies that work for them

Presentation

Modeling: Teacher demonstrates new strategies, emphasizes their value and draws up a checklist of strategies for subsequent use

Research Methodology

3. Teacher provides opportunities for practice

4. Teacher assists learners in evaluating their success with the new strategies

Learners are provided with opportunities to practice the new strategies Learners are shown how the strategies can be transferred to other tasks Learners are provided with further tasks and asked to make choices about which strategies they will use

Practice

Teacher helps learners to understand the success of their strategy use and assess their progress toward more self-directed learning

Evaluation

109

General practice. Learners are given a range of tasks to deploy new strategies

Expansion

Action planning. Learners are guided to select strategies that will help them address their particular difficulties. Further practice and fading out of reminders to use Evaluation. Teacher guides to evaluate progress and strategy use and to set themselves new goals

Among the models, Chamot et al.’s (1999) model was chosen for the present study. The model has been used successfully by second language teachers. The key organizing principle is centered around the students’ mental processes for solving problems. There are five main phases in the model (see Figure 4.1), which have been used in most models of SBI (Macaro, 2009).

Chapter Four

110

Teacher Responsibility Preparation Activate Background Knowledge

Presentation Explain Model

Attend Participate

Practice Prompt Strategies Give Feedback

Evaluation Assess Strategies

Expansion Support Transfer

Apply Strategies with Guidance Assess Strategies Use Strategies Independently

Transfer Strategies to New Tasks

Student Responsibility Figure 4.1. A strategy-based instruction model (Chamot et al., 1999, p. 46)

The model begins with a “preparation” phase in which learners’ current learning strategy use on a familiar task is discussed. In the “presentation” phase, the teacher explains and models the target strategy. The teacher also asks learners if and how they have used the strategy before. The third phase is “practice”. Learners need to use the new strategy in a task. The “evaluation” phase requires learners to evaluate their strategy use immediately after their strategy use. The final phase is “expansion”. Learners are to transfer the new strategy to a new task. The five recursive phases allow the teacher to shift among phases catering to learners’ needs for understanding and using the strategies. As was made clear in chapter 3, the concepts of ZPD and scaffolding (Vygotsky, 1978, 1986) were also incorporated as important theoretical underpinnings to inform the present study. This model well fits the conceptualization of ZPD and scaffolding. As shown in Figure 4.1, the teacher assumes most of the responsibility in the first phase, “preparation”. In this phase, learners’ role is to “attend and participate”. In other words, they are mostly other regulated. Gradually, the teacher releases some responsibility to learners, who are expected to do some self-regulation. In the “expansion” phase, learners are expected to self-regulate themselves and assume most of the responsibility. To be specific, they are expected to transfer strategies to new tasks and use their strategies independently. In addition, substantial instructional information and accompanying rationale have been provided in Chamot et al. (1999)

Research Methodology

111

for the present study to refer to, as compared with the other models. Therefore, Chamot et al.’s (1999) model was adopted as an overall guide, though with some variation for the intervention. Principles of MLE in Lesson Plans and the socio-cultural dimension of LLS As mentioned in 3.2. and 3.3, the socio-cultural dimension of LLS is actualized through the mediated learning experience between the teacher and learners, situated in a multi-dimensional view of language learning. Therefore, the lesson plans were designed following three important conditions for MLE. The three criteria are intentionality, mediation of meaning, and mediation of transcendence. All the three principles were embedded in the design of the lesson plans. In achieving intentionality, for example, the teacher in the classroom should be more focused on what the strategies were to solve the language problems, and how the students could apply the strategies to the task at hand with the resources available to enhance their writing. Take planning strategies for example: The teacher should explain what planning strategies were, what were involved, and how planning strategies should be used. Simply put, the thinking processes of the teacher should be either explicitly or implicitly shown to learners through classroom interactions because the teacher, as a more successful writer or learner, could demonstrate to the students how LLS worked in context. That is also why thinking aloud has been employed as a useful method in most studies of strategy-based instruction (e.g., Chamot et al., 1999; Cohen, 1998; Grenfell & Harris, 1999). With the intention of mediating on learners’ use of LLS for a task, the teacher (mediator) could help modify learners’ cognitive processes, e.g., language learning strategies. Mediation of meaning was realized as follows. After the students performed a writing task using the strategies introduced by the teacher, the teacher would help the students realize what strategies they had employed completing the task. Such questions as “what strategies did you use?”, “were they useful?”, and “how did you use them?” were used to enable the students to see what had been achieved and how the achievements had been made. The third condition, transcendence, was to help the students transfer lessons learned from one experience to another. For instance, goal setting in lesson one was for the students to see the importance of having writing goals for a particular writing task. In this lesson, learners were led to come to realize that they needed to set writing goals for any piece of

Chapter Four

112

writing through talking about the goals they set for their everyday life in sports or saving money. A few responses were elicited from the students and written on the whiteboard. Then, questions like “what kind of goals do you set for your English language learning?” were addressed to the students. The purpose was to show the students that goal setting was a useful strategy for many different tasks. Different goals may influence their efforts to learn in different ways. For example, the teacher could ask the students whether they set goals for themselves when they wrote a particular text and how this might influence their writing. Intervention piloting As there were not many SBI intervention studies of this nature, the research team decided to conduct a pilot study to examine how data collection instruments, research methods, and procedures could be improved. The research team planned to involve two classes in a primary school. However, upon the school’s request, 240 pupils from six intact primary 5 classes were included as the school would like more teachers to be involved in educational research. Three classes were assigned as experimental classes, and the other three as control classes. Care was taken to make sure that the experimental classes were comparable with the control ones. At this stage, all the possible data collection instruments that would be used in the intervention study were designed and piloted. Before the pilot, two briefings were conducted. One was for the experimental teachers, and the other the control teachers. In the briefing for the experimental teachers, five important topics relevant to the intervention were shared with them besides some general issues pertaining to the project. The topics included: 1. 2. 3. 4. 5.

What are learning strategies? What is SBI and why is it important? How are SBI lessons conducted in the classroom? What writing strategies do pupils use? How should the SBI lesson plans designed for the pilot be carried out?

The teachers’ questions and concerns were cleared in this briefing. In the briefing for the control teachers, they were informed of the purpose and procedures of the intervention. They were told to conduct their writing lessons as usual and not to find out what was happening in the experimental classes. Should the intervention be successful, relevant

Research Methodology

113

information on SBI would be provided to them afterwards. The pilot lasted 9 weeks and all the sessions by the experimental teachers were observed and video-taped. At least, two sessions by each control teacher were observed and video-taped. As all the instruments and procedures in the pilot were also used in the final intervention, I will only mention those to which changes and improvements were made in light of the pilot. The time for each lesson was 30 minutes in the pilot. It turned out that all the three experimental teachers could not finish their teaching within the time frame in the first week of the pilot. They suggested that 45-60 minutes would allow them to do a better job. Therefore, from the second week onwards, the time was extended to 45-60 minutes. In terms of difficulty level of the content for each lesson plan, the teachers provided positive feedback. They found that the lessons were suitable for their P5 pupils. Even the weaker class in the experimental group could follow their teaching. This suggested that the SBI lesson plans fell within the P5 pupils’ ZPD. In the second teacher conferencing, a teacher suggested that more information on the Guidelines section in the lesson plans would furnish them with more knowledge to understand the project better. There were only two components, strategy objective and strategy rationale in this section. Therefore, I decided to provide more details in these components and add one more component, strategy description so that the teachers could be equipped with a better understanding of the SBI lessons. The research team also felt that more details should be added to the Procedures section in the lesson plans. Based on the feedback and suggestions mentioned above, I revised the lesson plans used in the pilot. Then, the lesson plans were scrutinized by a key member of the research team for improvement. I revised the lesson plans again on the basis of the feedback and suggestions from the key member. Intervention lesson plans and target strategies Four groups of target strategies, i.e., planning strategies, feedback handling strategies, revising strategies, and text-generating strategies (see Table 4.4), which were also included in the writing strategy questionnaire (Phase 1 questionnaire), were taught in nine lessons. The criteria for selection of the writing strategies for the SBI were as follows. First, a preliminary list of writing strategies that were found to be highly correlated with the pupils’ English language results was established. Many

Chapter Four

114

of these strategies were reported to be significantly correlated with writing proficiency, as reviewed in chapter 2 (e.g., Chamot & El-Dinary, 1999; Gu et al., 2005; Victori, 1999; Zamel, 1983). Weigle (2005) also summarized a series of similar writing strategies that skilled writers use. Then, the strategies that represent the process of writing (Flower & Hayes, 1981) were selected. For example, setting writing goals was found to be significantly correlated with good writing results and it is a very important strategy for students. Therefore, it was taught in the first lesson. Table 4.4. List of SBI Writing Lesson Plans Lesson # 1

Theme

Target strategies

Setting writing goals

Planning strategies

2

Planning for content

3

Assessing the audience

4

Getting ideas-- reading about the topic Attending to language at the word level

Planning strategies, textgenerating strategies Planning strategies, textgenerating strategies Planning strategies

5 6 7 8 9

Text-generating strategies, revising strategies

Attending to grammatical structures Writing the essay: orchestrating strategies Revising the essay

Text-generating strategies, revising strategies Text-generating strategies

Getting and responding to feedback

Feedback handling strategies, revising strategies

Revising strategies

It should be pointed out that each group of target strategies consists of several sub-strategies that work collectively to facilitate the writing process (e.g., Flower & Hayes, 1981). For example, planning is comprised of such general processes as setting goals, generating ideas, and organizing ideas into a writing plan (e.g., Flower & Hayes, 1981; Graham, 2006a, 2006b). Each lesson may involve two or more types of strategies as they were closely related in the writing process. For example, while the main focus of lesson 3 was planning strategies, some of the strategies taught in the lesson could also be used for revising.

Research Methodology

115

According to Martin (2008), there should be a test for the discriminant validity of the intervention by including some non-target variables (scales) to safeguard the intervention. The Writing Strategy Questionnaire used in Phase 1 included eight variables. The same questionnaire served as both the pre-intervention questionnaire and the post-intervention questionnaire for the intervention. Therefore, the other four variables, i.e., self-initiation, monitoring and evaluating, resourcing, and help-seeking and affectmanaging (other than the four focused groups of strategies in the lesson plans), were treated as non-target variables and would provide a measure to test if the intervention also yields positive results on them. This additional step helps to make sure it is the intervention per se rather than mere participation in a program that impacts positively on the target variables (strategies). As noted by Pressley, Mohan, Fingeret, Reffitt, and Raphael-Bogaert (2007), effective writing instruction should always entail some form of plan-draft-revise (see also Harris, Santangelo, & Graham, 2010 for a detailed review). In a similar vein, the nine lessons were designed based on the three important writing processes: planning (Lessons 1-4), writing/drafting (Lessons 5-7) and revising (Lessons 8-9) and structured as follows. Lesson 1 aimed to develop in pupils an awareness of the importance of writing with a clear purpose through the use of goal setting (e.g., Flower & Hayes, 1981; Graham, 2006a). It is very important for young writers to determine the purpose(s) of a particular writing task before they embark on the actual process of writing. Through goal setting, young writers can keep themselves on the right track during the writing process. Good writers constantly refer to their pre-defined goals and check if their writing has been effectively orchestrated towards achieving their purpose(s) during and after the composing process. In this way, their understanding of why they write something informs their decision concerning what to write and how to write it effectively. In the lesson, the pupils were first instructed to reflect on the goals they set in their everyday lives as well as the ones for their learning. Then, they were asked about the goals they set for English writing. This was supported by the teacher’s modeling for a writing task on a dangerous road accident through answering: What happened in the story? What damage occurred? Who were involved? Why did it happen? Where did it happen? How did it happen? They were also encouraged to note down their writing goals on a piece of paper before writing. Then some students were called upon to share their writing goals with the rest of the class. All the pupils were also

116

Chapter Four

encouraged to share their comments on how to make the writing goals more effective. Lesson 2 taught strategies of planning for content prior to starting to write, involving the three main elements for narrative writing, i.e., characters, setting, and event. Effective writers employ these strategies in order to have an interesting and well-developed story line. They either plan their content mentally or on paper before writing so that they can refer to their plan during the process of writing or revising. Effective writers also check their plan for content from time to time when writing. If they see the necessity to make changes to it, they will do so accordingly. However, most ineffective writers do not have a plan for their content at all. Often their characters are not developed and few details are presented about the setting. Some of them might lose track while they are writing without a plan. In this lesson, the teacher modeled how planning for content (a dinner experience at a hawker centre with family members) was carried out by asking the pupils the following explicit questions: How often does your family go to hawker centers for dinner? How many members are there in your family? Which hawker center do you usually go to? What does it look like? Is it crowded? What does your family usually order? Has anything unusual ever happened while you were there? What was it? Did the unusual event interrupt your dinner? Did your family members become involved in the incident? What did they do? What did you do? These content-related questions help writers plan for their content before writing. Some pupils were asked to respond o these questions and the rest were invited to provide comments. The teacher then provided feedback on the extent to which they had used the strategies. All the pupils were then required to plan for their own content on paper and share their plan with their friends later. In Lesson 3, the pupils were taught how to assess their audience. Stories intended for different audiences present information and use language differently. When writing, writers need to assess who the audience is, what kind of background knowledge they have, and what they expect to know from their stories. The pupils were required to talk about (with teacher modeling) what happened at a hawker centre for two different people: their grandmother (who often visits them and goes to the hawker center with them) and a stranger (who does not know them at all and has never been to the hawker center before). The pupils were asked to respond to questions such as Do you need to tell your grandmother/ the stranger about your family members? Why (not)? Do you need to tell your

Research Methodology

117

grandmother/ the stranger about the hawker center? Why (not)? The teacher would provide feedback to the pupils’ responses to the questions afterwards. Once they became familiar with this strategy, the teacher then asked the pupils what kind of language they would use (simple or advanced) when they write for another two different audiences: young kids and classmates, and whether their story would be long (for classmates) or short (for young kids). The teacher would then provide feedback. These activities helped the pupils realize that they should write the same story in a different way if the audiences are different. Lesson 4 focused on strategies for getting ideas. The teacher explicitly suggested that the pupils should read others’ stories and essays to improve on their own writing. Getting ideas by reading is an important writing strategy that all effective writers know and use. Getting ideas does not only mean to learn what other writers’ plots (stories) are like; effective writers also learn how others present their information and how to use language in their texts. In this lesson, the pupils were asked to read two stories about a dinner at a hawker centre. Then, they were asked to make a comparison between the two stories and decide which one was better and why. For example, the pupils were prompted with questions like Which story do you think is better and why? Is it because of the way that the characters or settings that are described in the story? Or is it because of what happens or the ending of the story? They were told to be sure to have a valid reason for why one story is better than the other. Then, the teacher asked the pupils to identify useful information that they could use in writing their own stories. A few pupils were called upon to share their answers with the class. Again, the teacher raised the pupils’ awareness on getting ideas by reading about a topic. Lesson 5 was to teach the pupils to how to choose words appropriately when writing a composition. Expression and clarity of meaning can be enhanced by giving attention to language at the local level, e.g., appropriate vocabulary, sentence structures, and cohesive devices. The strategy was to help young writers become more aware of the language resources at their disposal and how this strategy can help them improve their writing. Firstly, the teacher pointed out to the pupils that there were many English words that are similar in meaning but differ in subtle ways. They were given a list of verbs, i.e., reply, chant, comment, shout, grumble, answer, shriek, remark and advise to discuss the similarities and differences between them. This activity was to raise the pupils’ awareness of how effective word choice can help enhance the subtlety in meaning

118

Chapter Four

and quality of writing. Then, the teacher provided the following two statements: “Chandra advised that you should leave” versus “Chandra said that you should leave” to help the pupils realize how word choices would give the reader a better picture or sense of what the character was doing. The next activity the pupils were asked to do was to critique the vocabulary used in a story about a dinner at a hawker centre. They had to distinguish the vocabulary that described accurately what the character was doing from those that were not descriptive enough. The less descriptive vocabulary was then replaced with more descriptive ones by the pupils. At the end of the lesson, the teacher raised the pupils’ awareness again on how the strategy, attention to language at the word level, could help them to improve their writing by reiterating the key points in this lesson. Lesson 6 taught strategies for attending to grammatical structures in story writing. Effective writing depends not only on what is said but also how it is said. Infelicitous grammatical structures can detract from a good story. In the context of narration, the verb tense is also an important area in the writing of stories. While the past tense is commonly used in story writing, other tenses are also possible depending, among other things, on the specific relationships between the event times and the time of reporting. Therefore, attending to grammatical structures in story writing is an important writing strategy that pupils must learn and employ in writing stories. In this lesson, the pupils were given four openings to the same story with different sentence structures and tenses. They were then asked to discuss how they were different and what contributed to the differences and the effectiveness of the openings. The teachers were specifically reminded that this lesson was about the different effects that different grammatical structures and tenses could create; it was not a grammar lesson. Then, the pupils were given two stories in which there were inappropriate grammatical structures and tenses for them to improve on, thus allowing the pupils to see the importance and effects of attending to grammatical structures in writing stories. In Lesson 7, the pupils were to apply the target strategies that they learned in the first six lessons to a specific writing task and to practice orchestration of those strategies. The writing task required the pupils to write a composition of at least 150 words describing what happened at a hawker centre while they were having dinner. The teacher and pupils worked out a list of considerations based on the writing strategies covered in the previous lessons so that the pupils were able to recall and internalize

Research Methodology

119

those writing strategies. In this lesson, the pupils were to orchestrate their strategy use. To facilitate the process, the pupils were asked to write their writing strategy list on a piece of paper for their reference later. When the pupils failed to recall certain strategies, the teacher thought aloud and elicited those strategies without directly telling the pupils what the writing strategies were. They were reminded that they might need to employ a group of strategies that they learned in different lessons at the same time. Although the writing strategies were taught sequentially in the previous lessons, some of them could be used for the same problems/tasks. For example, goal setting could be orchestrated with other strategies, such as planning for content, assessing the audience, and generating ideas. Then, the pupils were given about 40 minutes to complete the writing task. The teacher reminded the pupils of employing the writing strategies on their list at a few times during the process of completing the writing task. The writing strategies taught in Lesson 8 were revising strategies. The pupils were asked to revise the story that they wrote in the previous lesson. Revising entailed a group of strategies concerning different aspects, such as content, organization, and language. The pupils were reminded to check all the aspects. The teacher then guided the pupils to design a checklist for revising which consisted of the three general aspects/categories, i.e., content, organization, and language. Under each category, the pupils were encouraged to come up with as many points as possible. The teacher’s role was to facilitate the process rather than providing answers to the pupils. The pupils were then asked to revise their own compositions according to the checklist. During their revising process, the teacher encouraged them to follow their checklist as much as possible. At the end of the lesson, individual pupils were called upon to reflect their revising process. They were also asked to make comparisons between their revising in the past and their current revising process for them to see the benefits of employing the target strategies in this lesson. Lesson 9 aimed to teach the pupils strategies for getting and responding to peer feedback. Research has shown that peer feedback can enhance revision (e.g., Hu, 2005; Min, 2006; for a detailed review, see Fitzgerald, 1987). Therefore, the revising process should also involve eliciting suggestions and comments from peers with regard to content, organization, and language. Effective writers often get feedback on their writing from their peers and other sources in order to help themselves revise their writing and make it more effective. First, peer feedback can create a context in which children can participate actively in the revision of their

120

Chapter Four

writing. Second, it provides an opportunity for other-mediated learning of useful revising strategies, a crucial precursor to independent strategic use of these strategies. Third, it raises awareness in children that writing is not an individual but a social activity and consequently, socializes children into a real-world practice. In addition, this strategy provides young writers with an audience of readers who give feedback to their writing, identify strengths and problems, and make suggestions for improvements, “a social nudge to consider audience perspectives on their writing” (Rowe, 2008, p. 410). Young writers can also learn from working as both author and editor (MacArthur, 2007). In the lesson, the teacher firstly told the pupils about the benefits of using feedback handling strategies. Then, the teacher demonstrated through thinking aloud how to provide polite, constructive and helpful suggestions and comments on the story that the pupils wrote in the previous lesson. For example, the pupils were reminded “What suggestions can I, the reader, give the writer?” (Sengupta, 2000, p. 103). The pupils were taught to decide whether they would like to follow their peers’ feedback or not with valid reasons. At the end of the lesson, the teacher raised the pupils’ awareness again on the benefits of employing the target strategies by asking the pupils if they found feedback handling strategies useful and why. The pupils were also asked if they were open enough to give their peers feedback or comments and why. On the whole, a central purpose of the nine lessons was to teach students the target writing strategies and help them understand the complexity of the writing process (Pritchard & Honeycutt, 2007). There were two major sections in each of the nine lesson plans introduced above. The first section presented the guidelines for teachers containing the following: language objectives, target strategy (strategy objective), strategy description, strategy rationale, time needed, and materials. This section was to provide the teachers with a solid rationale and the necessary information on the materials. As mentioned previously, the lesson plans were designed following Chamot et al.’s (1999) model with some adaptations. The adaptations resulted from the feedback and suggestions collected in the pilot as well as my own observations. The second section of each lesson consisted of the following procedures: review of the previous lesson, preparation, presentation, practice, evaluation, and summary. The review section was for the teacher to help the pupils to reflect on how the target strategies learned in the previous lesson facilitated their writing. Questions about the usefulness and implementation of the strategies learned in the previous

Research Methodology

121

lesson were asked. Also it was made clear that the strategies in the current lesson were well connected with the previous strategies during the writing process, together helping improve their writing. The preparation stage required the pupils to think actively what writing strategies they employed for their writing in relation to a specific writing task. It was also for the pupils to warm up and activate their background knowledge. This stage was for the teacher to raise the pupils’ awareness of their strategy use. In the presentation stage, the teacher introduced the writing strategies to be taught in that lesson. This stage was to make the pupils understand the following points: “What are the target strategies?”, “When are they normally used?”, and “Why are they used?” During this part of the lesson, the teacher was required to do frequent think-alouds. The purpose for the teacher to think aloud was to show the pupils how, why, and when writing strategies were used for writing. The practice stage was for the pupils to apply the target writing strategies to a particular writing task. At the evaluation stage, the pupils were asked to reflect upon the writing strategies they just learned and used. For example, the pupils were asked to reflect on the use, usefulness, and contexts of target strategy use in relation to a specific writing task at the practice stage. The final stage, summary, was to ask the pupils to sum up what was covered in the writing strategy lesson. This stage was to raise the pupils’ awareness again of the target strategies learned in the lesson for the purpose of reinforcement. The teacher was also encouraged to provide the pupils with ample opportunities to apply the target strategies in other English lessons for the week. A timeframe was suggested for each stage. The time needed for each stage may slightly vary depending on the actual need in the class. In addition, the time suggested for each stage was only a general guideline. However, the total time used must not exceed 60 minutes. The experimental teachers were allowed to vary the time where necessary. It should be noted that although the pupils in the experimental group only practiced writing about the same story during the intervention, they carried out a variety of tasks that were closely related to the topic. The pupils in the control group wrote about three to four compositions on average during the period. PowerPoint slides for teachers were not originally included in the lesson plans. However, two teachers from one school expressed their interest in designing their own PowerPoint slides based on the lesson plans given to them. The research team also felt that the SBI would be more effective if the experimental teachers could design the slides by themselves. Through designing their own slides, their understanding of the lesson plans could

122

Chapter Four

be deepened. Therefore, the research team decided to convince the rest of the experimental teachers of the potential benefits. In the end, all the six experimental teachers agreed to design their own instruction slides with my assistance during the teacher conferencing sessions prior to and after each lesson. In addition, the experimental teachers in the same school shared their slides so that common understanding of each lesson could be reached.

4.2.3 Instruments Besides the lesson plans outlined in the previous section, the instruments used for the intervention phase include pre- and post-intervention questionnaires, pre-, post-, and delayed writing tests, semi-structured interviews, class observations, and teacher logs. Various data collection methods as mentioned above were employed in the intervention phase. The purpose of the multiple methods was to achieve triangulation (Mizumoto & Takeuchi, 2009; for detailed reviews, see Dörnyei, 2007; Mackey & Gass, 2005). For example, questionnaires may not fully capture the complex interactions between strategy use and language learning achievements. Multiple data collection methods (both quantitative and qualitative research methods) can complement each other. Mackey and Gass (2005) note that triangulation helps explore research issues from all feasible perspectives. Another important purpose is to reduce the researcher bias and enhance the validity and reliability of the information collected and analyzed (Johnson, 1992). Pre- and post-intervention questionnaires The writing strategy questionnaire used during Phase 1 was administered both as a pre- and post-intervention questionnaire. As there were nine weeks in between the pre- and post-intervention questionnaire administration, although the same questionnaire was used twice, the pupils would not remember their responses to the pre-intervention questionnaire. All the participants were informed that they should respond to the writing strategy questionnaire according to what they thought; there were no right or wrong answers. Their responses would not in any way affect their English language results. The purpose of using the same questionnaire in the postintervention administration was to examine differences in students’ writing strategy use after the SBI. The internal consistency reliabilities (see Table 4.5) of both the pre- and post-intervention questionnaire as measured by

Research Methodology

123

Cronbach’s alpha (Į) were acceptable (see DeVellis 1991; Petriü & Czárl, 2003). Table 4.5. Internal Consistency Reliabilities of the Pre- and Postintervention Questionnaires Strategy

Self-initiation Planning Monitoring & evaluating Text-generating Revising Feedback handling Resourcing Help-seeking & affectmanaging

Pre-intervention questionnaire Scale alpha .662 .761 .656

Post-intervention questionnaire Scale alpha .708 .730 .608

.706 .699 .554 .641 .588

.743 .694 .596 .623 .589

Pre-, post-, and delayed tests There were no national standardized writing tests available at this level for the pupils. The research team therefore had to design our own tests for the larger research project of which this study was a part. Three parallel writing tasks were designed for the pre-, post-, and delayed tests to alleviate instrument-related effects. The tasks were narratives, a genre required in the syllabus for primary 5 pupils in Singapore primary schools. Two experienced primary school teachers were consulted to evaluate the topics. They both found the topics suitable for primary 5 pupils. The three tasks required the pupils to write a 150-word story about what happened based on a short description given. The first task (A) was about a lost wallet. The pupils were asked to write a story on what happened after they found a lost wallet containing money and an IC. For the second task (B), the pupils were instructed to describe what happened on a morning when they had a terrible stomachache. The third task (C) required the pupils to write a story on helping an elderly person who was shouting for help. The purpose of these tests was to measure writing proficiency change for both the experimental and the control classes at the end of the intervention. Each task took about 40 minutes for the pupils to complete. In order to

124

Chapter Four

ensure comparability of the three tasks, a pilot was carried out at another local school in which 66 pupils took the tests. They were randomly assigned to three groups. The first group (AB) of pupils took both task A and task B. The second group (BC) took task B and task C. The third group (AC) took task A and task C. The Pearson correlation between the two tasks was .687 for the AB group, .774 for the BC group, and .854 for the AC group. All were statistically significant (p < 0.05). It seemed that the correlation for the AB group was not as good as the other two. I plotted the results for the AB group using SPSS, which showed an obvious outlier. One pupil obtained 46.5 for composition B, however, only 26 for composition A. This showed that the student did not perform consistently on the two tasks, which caused the vast difference between the two scores. The form teacher of the student reported that the student did not complete composition A due to sickness. Therefore, I removed this pupil’s scores and performed the Pearson correlation test for the AB group again. The correlation for the AB group was .784 well within the acceptability range. In addition, paired samples t test did not show significant differences between task A and task B, task B and task C, and task A and task C, which means that each pupil scored similarly for the two particular tests completed by him/her. Therefore, the above analysis showed that the three tests were comparable in terms of difficulty. They can be used to measure pupils’ writing proficiency. Before the intervention, a pre-test was conducted for all the pupils. Each class was equally divided into 3 random groups. The pupils in both the experimental and the control groups would stay in the same group for the tests throughout the whole intervention period. The teachers were given a grouping sheet so that each pupil would take all three writing tests in their own respective group. The tests were conducted as follows. In the pre-test, group one was assigned to complete task A, group two task B, and group three task C. In the post-test, the topics were rotated to a different group. Group one completed task B, group two task C, and group three task A. Likewise, group one did task C, group two task A, and group three task B in the delayed test. In this way, each participant took one different writing task for each test at a different time point while the three tasks were counterbalanced for each group. The purpose of rotating the tasks was to minimize instrument-related effects and to avoid the problem of having the pupils write about the same topic three times. The three tests were marked according to a composition marking scheme, which was adapted from the rubrics of the Singapore Primary School

Research Methodology

125

Leaving Examination (PSLE) English language Continuous Writing Marking Scheme. This is an established marking scheme for primary school composition marking from the Ministry of Education (MOE) Singapore. There are three main categories, i.e., content, language, and organization, on the PSLE Marking Scheme. Singapore primary school teachers are very familiar with it as they are assigned to PSLE marking every year. The three categories are valid aspects to attend to in assessing pupils’ writing proficiency (Weigle, 2002, 2007). However, the research team decided to break “language” down into two subcategories, i.e., grammar and spelling, and vocabulary. We felt that by including such more refined categories, it could better facilitate the raters’ marking. The adaptation was also based on the feedback from the raters. They mentioned that the category “language” was a bit too general for most of the teachers. It would be better to have more detailed categories. Therefore, there were four components: content, grammar and spelling, vocabulary, and organization, with each component ranging from 1 to 20 marks on the marking scheme for the present study. Two primary school teachers who were involved in PSLE composition marking were assigned to mark all the compositions independently. Before they started marking the three tests, a series of training sessions were held in order to establish inter-rater reliability. First, a meeting was held to discuss the marking scheme to ensure they had similar understandings of the scheme. Then, they were asked to mark six compositions independently. A few discrepancies were found between the two raters. They then discussed their differences with reference to the marking scheme until agreement was reached. Lastly, they were given 28 compositions collected from the piloting of the writing tests to mark. Their inter-rater reliability was 0.85 (Pearson’s r), which allowed me to conclude that their marking was well within the range of acceptability. The information on the pupils’ compositions that would allow the teachers to identify whether a particular composition was from the pre-, post-, or delayed test was covered. Then, they were asked to mark all the compositions independently. Therefore, two sets of marks for each pupil were generated. Both raters’ scores were added up and averaged for each pupil. Teacher interviews Both pre- and post-intervention teacher interviews were conducted for both the experimental and the control group teachers. The pre- and post-

126

Chapter Four

intervention teacher interviews were designed for eliciting the teachers’ perceptions and behaviors with respect to writing strategies. In each interview, both prepared and spontaneous questions were asked to the teachers. The structured interview was for the teachers to answer such questions as their age, years of teaching experience and so on. The semistructured interview allowed for flexibility in eliciting their responses where there was a need to probe more. It covered three main aspects, i.e., background, teacher perceptions, and teacher behaviors. The purpose of the teacher interviews was to find out if there were differences in their practice and understanding of writing strategies before and after the intervention, what their attitudes were towards the intervention, and what suggestions they had. Altogether, 12 teachers were interviewed. All the interviews were audio-recorded and transcribed. Student interviews In order to gain a better understanding of the pupils’ strategy use and their writing process, interviews were conducted with case study pupils from both the experimental and the control groups before and after the intervention. Questions on their perceptions of strategies, and attitudes and reactions towards the SBI were asked. Similar to the teacher interviews, both structured and semi-structured questions were asked, which consisted of three aspects: 1) background, 2) pupil perceptions, and 3) pupil behaviors. Altogether, 24 pupils from the experimental group and 24 from the control group were interviewed. All the interviews were audiorecorded and transcribed. Teacher journal logs Each experimental teacher was given a teacher journal log (Appendix 1) to reflect upon their instruction during the intervention period. The journal log data was collected to provide information about the experimental teachers’ changes in their beliefs and practices as well as their suggestions for the SBI during the whole intervention phase.

4.2.4 Procedures Prior to each lesson, the lesson plan was sent to the teachers. Then, a conferencing session was held between the teacher and the researcher on how the lesson was to be conducted. Each conferencing session lasted about one hour. The researcher and the experimental teachers would go

Research Methodology

127

through the lesson plan to ensure that the teachers understood the lesson and what they were expected to do. Such important concepts as ZPD, scaffolding, and MLE and associated ideas were discussed so that they could understand the theoretical grounding of the SBI and draw on the ideas (e.g., mediations) in their teaching. After each lesson, the researcher and the experimental teachers would meet again to reflect upon the issues, problems, and concerns arising from each lesson both on the teachers’ and the researcher’s parts. Another purpose of the conferencing was to ensure that all the teachers in the experimental group would administer the SBI faithfully. The intervention, which lasted 11 weeks including both the pre- and posttests, was implemented without major difficulties. The delayed test was administered one month after the post-test. Nine SBI writing lessons (nine weeks) were conducted for the experimental classes with one lesson in each week. As there were 54 sessions altogether for all the 6 experimental classes within the 9-week period, I was not able to observe all of them. However, I tried my best to schedule sessions with the experimental teachers together so that the number of observations could be maximized. In the end, 44 observations were conducted for the experimental classes using the observation log (Appendix 2). In a typical SBI session, I would sit at the back of the classroom, jot down my reflections on the lessons in the log and video-tape the session. As mentioned earlier, I would also talk to the teachers about the implementation of a lesson based on my observations during the weekly conferencing. The control teachers had their regular classes concurrently as their experimental counterparts. Immediately after the intervention, all the pupils from both experimental and the control groups completed the post-invention writing strategy questionnaire and the writing post-test. A delayed test was administered to them one month after the intervention.

4.2.5 Analyses Descriptive statistics of the two questionnaire administrations were generated using SPSS 21. The post-intervention questionnaire data were compared with the pre-intervention data to examine the differences in the pupils’ writing strategy use before and after the SBI. As the participants were not randomly selected for the SBI, there could be prior differences among the classes. The prior differences need to be partialled out in the analysis. Therefore, repeated measures one-way analysis of covariance

128

Chapter Four

(ANCOVA) on the tests and questionnaire data was conducted. This was to examine the effects of strategy instruction on the pupils’ writing test results. All the interviews with both the pupils and teachers were transcribed and analyzed in order to reveal any qualitative differences before and after the intervention for the case study pupils. The interview data served to further strengthen the analysis of the change in the pupils’ strategy use after the intervention. The teacher journal logs and observation journal logs were also analyzed. The data collected from these two sources provided additional information for a deeper understanding of the pupils’ LLS use and the effects of the SBI.

4.3 Ethical considerations I touched upon the ethical considerations of each phase in this study as I went along. This section will be summarizing the major considerations and reiterating some of the most important ethical considerations. Strict adherence to ethical standards in two phases of the project was most necessary and important for me to consider during the process of the research. Ary, Jacobs, Razavieh and Sorensen (2006) state that education researchers have obligations both to their human subjects and their profession. I also consulted the Code of Ethics of American Educational Research Association (2011) to note some of the possible ethical issues and solutions for this study. Before the research started, all the prospective participants, e.g., the students and their guardians, and their teachers, were informed of the purpose of the project and what would be required of them. The participants’ informed consent was obtained. They were assured that they had the right to withdraw from the study at any time without any effect on their studies and status in their schools. At the same time, they were informed that the confidentiality of their information or data would be ensured. Their anonymity would not be compromised at any time before, during and after the project. During the process of the research, I took every necessary effort to make sure that the participants were treated equally and with respect. Their questions and problems that arose due to the project were duly attended to. Clear guidelines and instructions were made available to the participants so that they understood what was expected of them.

Research Methodology

129

The data collected from the participants were treated with great care so that their identities would not be revealed. After the results were analyzed, I also communicated our findings and practical significance of the project. Detailed written feedback was given to each participating school/class. I was also obligated to use the data to increase the effectiveness of classroom teaching and benefit the students as well as the teachers.

CHAPTER FIVE FINDINGS AND DISCUSSION OF PHASE ONE

This chapter will report the findings from Phase 1 of the study, the writing strategy questionnaire survey. This phase was to identify the range and types of writing strategies employed by upper primary school pupils in Singapore, addressing the following two research questions: 1. 2.

What writing strategies do upper primary school pupils use? How are the differences in strategy use related to language competence?

5.1 Descriptive statistics for overall strategy use In this section, I will report descriptive statistics to note the overall pattern of writing strategies used by the participants. In 5.2, I will analyze the relationship between English language proficiency and reported strategy use based on the one-way ANOVA results.

5.1.1 Metacognitive strategies Table 5.1 shows the descriptive statistics for the 22 metacognitive strategies in three subscales, i.e., self-initiation, planning, and monitoring and evaluating. The Cronbach’s Į for each subscale is included in the parentheses. In distinguishing the frequency of strategy use, the present study employs Oxford’s (1990) categorization. A mean in the range of 3.5 – 5.0 is considered to be high frequency use, 2.5 – 3.4 medium frequency, and 1.0 – 2.4 low frequency strategy use. As can been seen from the mean subscale scores, the pupils reported using the three groups of strategies at a medium frequency (M = 3.05 for self-initiation, M = 3.37 for planning, and M = 3.17 for monitoring & evaluating). The most frequently used group of metacognitive strategies were planning strategies, whereas the least frequently used were self-initiation strategies.

Findings and Discussion of Phase One

131

Table 5.1. Descriptive Statistics of Metacognitive Strategies Subscale Selfinitiation (Į =.62)

Planning (Į =.75)

Strategy 1. Trying to find out how to write good English compositions. 2. Writing English compositions at home to improve writing. 3. Studying good English compositions in order to write well. 4. Asking for writing tuition to improve English writing. 5. Asking others about what they think about own writing and why. SUBSCALE 6. Discussing the topic with others before writing. 7. Reading about the topic before writing. 8. Writing out a plan before writing. 9. Thinking about what ideas to put down when planning. 10. Thinking about what words, phrases, and sentences to use when planning. 11. Keeping in mind teacher’s requirements for the composition when planning. 12. Thinking about how readers will feel about own composition when planning. 13. Thinking about how to organize ideas when planning. 14. Keeping in mind what type of composition to write when planning. 15. Planning a composition in mind before writing. SUBSCALE

n 3,177

M 3.65

SD 1.20

3,142

2.70

1.28

3,165

3.56

1.14

3,172

2.47

1.39

3,157

2.87

1.27

3,178 3,146

3.05 2.69

.79 1.27

3,160

3.83

1.16

3,151 3,173

3.13 3.86

1.26 1.09

3,151

3.70

1.16

3,140

3.39

1.17

3,161

3.15

1.25

3,170

3.25

1.23

3,160

3.56

1.13

3,167

3.18

1.33

3,175

3.37

.67

132

Monitoring & Evaluating (Į =.63)

Chapter Five

16. Sticking to what has been planned when writing. 17. Reading own composition aloud to look for mistakes. 18. Paying attention to spellings and grammar when writing. 19. Checking whether own composition meets teacher’s requirements after writing. 20. Comparing own composition with friends’ writing on the same topic. 21. Evaluating whether own composition is good or bad when reading it. 22. Anticipating whether readers will like own composition or not when reading it. SUBSCALE

3,134

3.36

1.18

3,143

2.27

1.27

3,156

4.00

1.12

3,148

3.21

1.21

3,148

2.82

1.26

3,149

3.14

1.19

3,159

2.90

1.22

3,175

3.17

.67

The individual strategies which fell into high frequency use included the monitoring and evaluating strategy of “paying attention to spellings and grammar when writing” (M = 4.00), the planning strategies of “thinking about ideas to put down when writing” (M = 3.86), “reading about the topic before writing” (M = 3.83), “thinking about what words, phrases, and sentences to use when planning” (M = 3.70), and “keeping in mind what type of composition to write when planning, and the self-initiation strategies of “trying to find out how to write good English compositions” (M = 3.65) and “studying good English compositions in order to write well” (M = 3.56). These strategies made up 32% of the metacognitive strategies. Two strategies (9%), “asking for writing tuition to improve English writing” (M = 2.47) and “reading own composition aloud to look for mistakes” (M = 2.27) belonged to low frequency use. The rest (59%) of the individual strategies in the metacognitive category fell into the medium range of frequency. The analyses of the metacognitive strategies suggested that the most frequently used group of strategies was planning strategies (M = 3.37). This group included a wide range of writing strategies, e.g., “thinking about what words, phrases, and sentences to use when planning”, “thinking about what ideas to put down when planning”, and “thinking about how to organize ideas when planning”. Such a finding corroborates

Findings and Discussion of Phase One

133

the literature in that young writers perceive planning as a critical element of good writing (see Graham & Harris, 2007). Both the high proficiency and the low proficiency pupils reported using planning strategies very often. Chamot and El-Dinary (1999) also found that planning strategies were most frequently by both the high proficiency and the low proficiency pupils. The least frequently used type was self-initiation strategies (M = 3.05), which consisted of such strategies as “trying to find out how to write good English compositions”, “writing English compositions at home to improve writing”, and “asking others about what they think about own writing and why”. This group of writing strategies mostly represents strategies that pupils use outside the classroom. The reason why the pupils used them least often may lie in the fact that young pupils only prefer to use strategies that their writing teachers teach them to use in class. In other words, seldom do they initiate their efforts in learning to write by themselves. Among all the metacognitive strategies, 68% of the strategies were used by the pupils from a low to medium frequency. Only 32% of the metacognitive strategies (seven metacognitive strategies) were employed at a high frequency. The findings may suggest that the pupils do use metacognitive strategies when writing. However, they do not use most of these strategies frequently. Among the most favored strategies, 57% were planning strategies (metacognitive strategies 7, 9, 10 & 14). When planning, primary school pupils thought about what ideas, words, phrases and sentences to put down in their text. They also read about the topic and considered the type of composition they would write. This finding indicates that Singapore primary school pupils engaged in planning more often than other strategies in writing. As was evident in my observations, although the writing instruction in the participating schools was mainly product-oriented, the writing teachers did teach the pupils how to plan sometimes, which may have resulted in the pupils’ preference for these planning strategies, as compared with the others. Two self-initiation strategies, i.e., “trying to find out how to write good English compositions” (metacognitive strategy 1) and “studying good English compositions in order to write well” (metacognitive strategy 3) also fell into high frequency use. While self-initiation strategies, as a group, were used at a medium frequency, it is understandable that these two strategies were favored by Singapore primary school pupils. In

134

Chapter Five

Singapore, it is a prevalent phenomenon that pupils buy composition books and read model compositions in order to improve their writing. Many primary school pupils generally study model compositions in order to either help them complete their writing assignments or brush up their writing skills in an ad hoc manner before the exam. The most frequently used individual strategy was “paying attention to spellings and grammar when writing” (metacognitive strategy 18). The finding suggests that Singapore primary school pupils pay great attention to mechanics and grammar when writing. As has been theorized in the literature of process writing, paying too much attention to surface issues, e.g., spelling and grammar, may hinder writers’ writing process. However, it is not surprising that the primary school pupils reported using this strategy very frequently in that this is a strategy that pupils can employ very easily. In addition, this finding may reflect the fact that many writing teachers focus too much on spelling and grammar when they teach writing and mark their pupils’ compositions. The two strategies that were least used included “asking for writing tuition to improve English writing” (metacognitive strategy 5) and “reading own composition aloud to look for mistakes” (metacognitive strategy 17). Although most pupils in Singapore attend tuition, many are often actually “volunteered” by their parents to such lessons. Attending private lessons may be the most popular activity that parents want their children to do in their spare time. It is rare that pupils would ask for tuition. It is also understandable that Singapore primary school pupils did not prefer “reading own composition aloud to look for mistakes”. For many pupils, writing is a silent process where they monitor their writing process silently.

5.1.2 Cognitive strategies Table 5.2 displays the descriptive statistics for the 18 cognitive strategies included in four groups: feedback handling (M = 3.20), revising (M = 3.16), text-generating (M = 3.13), and resourcing (M = 3.12). The average frequency of strategy use for the four groups fell into the medium range (2.5 – 3.4). Only one strategy “checking mistakes after getting back composition from teacher and trying to learn from them” (M = 3.54) was reportedly used at a high frequency. The use of the other individual strategies all fell into the range of medium frequency.

Findings and Discussion of Phase One

135

The most frequently used group of cognitive strategies was feedback handling (M = 3.20), which were comprised of “trying to remember or write down other people’s suggestions”, “thinking carefully about other people’s suggestions for own compositions”, “asking why if others ask oneself to make changes to own compositions” and “checking mistakes after getting back compositions from teacher and trying to learn from them”. This finding suggests that the pupils are used to handling feedback from other people, e.g., teachers, family members, and friends. They generally asked for help or tried to remember the suggestions from their teachers and family members. In many schools, it is a general practice that writing teachers provide conferencing sessions where students get feedback and suggestions on their writing. Many students may also ask for help from their siblings. The least used group of cognitive strategies were resourcing strategies, including “making use of dictionary to deal with lexical difficulties during writing”, “making use of grammar books, textbooks, or writing guides when having difficulty in writing”, and “revisiting old compositions for useful words, phrases, ideas when writing a new one”. Such a finding is not surprising as when writing, young children seldom consult a dictionary, grammar books, textbooks, writing guides or their old compositions. It is not a common practice among Singapore pupils to check the dictionary for the vocabulary to put in their compositions. They tend to use the words they are very familiar with. Many of them perceive writing as a non-stop process where they focus on producing text until they complete it. In general, young children do not refer to external resources while writing. Table 5.2. Descriptive Statistics of Cognitive Strategies Subscale Feedback handling (Į =.53)

Strategy 1. Trying to remember or write down other people’s suggestions for future use. 2. Thinking carefully about other people’s suggestions for own compositions. 3. Asking why if others ask oneself to make changes to own composition. 4. Checking mistakes after getting back composition from teacher and trying to learn from them. SUBSCALE

n M SD 3,139 3.07 1.25

3,140 3.08 1.18

3,133 3.13 1.31 3,150 3.54 1.18

3,175 3.20

79

Chapter Five

136

Revising (Į =.69)

5. 6. 7. 8. 9.

Textgenerating (Į =.72)

10.

11.

12.

13.

14.

15.

Resourcing (Į =.55)

16. 17.

18.

Changing spellings or punctuations when checking own composition. Changing words or phrases when checking own composition. Changing grammar when checking own composition. Changing ideas in own composition when checking it. (6) Re-organizing own composition when checking it. SUBSCALE Reading segments of own text to help oneself think of new ideas when writing. Reading other people’s writings for language to use in own composition during writing. Reading other people’s writings for ideas to write down in own composition during writing. Recalling ideas read elsewhere for use in own composition during writing. Recalling language from sources for use in own composition during writing. Re-reading teacher’s requirements to help oneself think of new ideas during writing. SUBSCALE Making use of dictionary to deal with lexical difficulties during writing. Making use of grammar books, textbooks or writing guides when having difficulty in writing. Revisiting old compositions for useful words, phrases, ideas when writing a new one. SUBSCALE

3,171 3.45 1.17 3,149 3.23 1.13 3,169 3.15 1.14 3,156 3.02 1.14 3,157 2.97 1.18 3,177 3.16 .77 3,152 3.35 1.17

3,464 2.81 1.29

3,163 2.76 1.26

3,151 3.29 1.21

3,147 3.37 1.17

3,134 3.18 1.18

3,175 3.13 .79 3,125 3.25 1.22 3,139 3.07 1.22

3,174 3.04 1.29

3,177 3.12

.90

Findings and Discussion of Phase One

137

Among all the cognitive strategies, only “checking mistakes after getting back composition from teacher and trying to learn from them” (cognitive strategy 4) fell into high frequency use, accounting for 6% of all the individual strategies in the cognitive category. The rest (94%) fell into the range of medium frequency. It is understandable that this strategy is popular among Singapore primary school pupils, who attend to their mistakes marked by their teachers. In Singapore, pupils are used to following what their teachers would like them to do in studies (Loh, 2007). The cognitive strategy 5 of “changing spellings or punctuations when checking own composition” fell into the upper range of medium frequency (M = 3.45), in the proximity of high frequency. In contrast, revising strategies (M = 3.45), as a group, were not very popular among the pupils. This was probably because writing teachers might have emphasized too much on spelling and punctuation in pupils’ writing process. As a result, pupils pay a great deal of attention to their spelling and punctuation. Another reason may lie in the less demanding nature of employing such a strategy (cognitive strategy 5) on the pupils’ side. For example, “reorganizing own composition when checking it” (M = 2.97) may require a drastic change to one’s writing, which may prevent primary school pupils from employing it freely. Spotting problems in spelling and punctuation in their compositions may be the easiest task that primary school pupils find, in comparison with the other revising strategies.

5.1.3 Social/affective strategies Table 5.3 presents the descriptive statistics for the social/affective strategies. The mean subscale score for this group of strategies was M = 3.16, with the frequency of individual strategy use falling into a medium range. The most frequently used social-affective strategy was “telling oneself to enjoy writing” (M = 3.47) and the least favored strategy was “rewarding oneself for completing an English composition” (M = 2.93). Another most frequently used strategy was strategy 5, “telling oneself not to worry when writing an English composition” (M = 3.45). The two strategies are affective strategies that can help learners deal with their stress in language learning. It was expected that the frequencies of use of the two strategies were close to high frequency in that writing was seen as the most demanding task by many primary school pupils in Singapore. Hence, it was not surprising that primary school pupils employed these two strategies to encourage themselves to complete their writing assignments

138

Chapter Five

due to the demanding nature of writing tasks. This finding is in accord with Loh (2007) who found that if the pupils were good in a subject, they would not use these strategies. If the pupils were not good in a subject, they would employ these strategies more frequently. Table 5.3. Descriptive Statistics of Social/Affective Strategies Subscale Helpseeking & affectmanaging (Į =.57)

Strategy 1. Asking teacher for help when having difficulty in writing. 2. Asking friends for help when having difficulty in writing. 3. Asking family members for help when having difficulty in writing. 4. Telling oneself to enjoy writing. 5. Telling oneself not to worry when writing an English composition. 6. Rewarding myself for completing an English composition. SUBSCALE

n 3,156

M SD 3.02 1.21

3,162

2.98

1.19

3,132

3.13

1.27

3,151 3,142

3.47 3.45

1.30 1.26

3,152

2.93

1.45

3,177

3.16

.72

It is interesting to note that “rewarding myself for completing an English composition” (social/affective strategy 6) was not really favored by the pupils. As shown in the qualitative data, the participants generally did not mention this strategy. The interview with the pupils in the questionnaire piloting also revealed they did not have the habit of rewarding themselves when they completed a composition. Generally, they might reward themselves or ask their parents to reward them if they did well in their exams. In summary, the pupils reported using a wide range of writing strategies at a medium frequency, consistent with the findings of the great majority of LLS descriptive studies (e.g., Lan & Oxford, 2003; Loh, 2007; Magogwe & Oliver, 2007; Wharton, 2000; Yang, 2007). A comparison of the three categories of strategies, i.e., metacognitive, cognitive and social/affective, shows that the pupils reported using metacognitive strategies (M = 3.23, SD = .60) more often than the other two categories (M = 3.15, SD = .64 for cognitive strategies and M = 3.16 SD = .72 for social/affective strategies), in support of the view that that metacognitive strategies play a critical role in writing (e.g., Harris, Santangelo, & Graham, 2010). The pupils were familiar with the need to manage their learning of writing, which entails

Findings and Discussion of Phase One

139

planning, monitoring and evaluating their own writing process, major elements of metacognition. Similar findings were also reported by HongNam and Leavell (2006), Nisbet et al. (2005), O’Malley and Chamot (1990), Padron and Waxman (1988), and Vandergrift (1997). These metacognitive strategies are thought to be especially potent mechanisms for developing competence in writing (for detailed reviews, see Harris et al., 2010; Zhang, 2010). Nonetheless, some studies have found different rankings among the different categories of strategies on their respective questionnaires. Wharton (2000) found social strategies were the most popular strategies among the undergraduates in Nanyang Technological University, Singapore, followed by compensation strategies, metacognitive strategies, cognitive strategies, memory strategies, and affective strategies on Oxford’s (1990) SILL. In Magogwe and Oliver’s study (2007), the primary school pupils’ most favored strategies were also social strategies, followed by metacognitive strategies, cognitive strategies, memory and compensation strategies. However, Lan and Oxford’s (2003) ranking was compensation strategies, affective strategies, metacognitive strategies, cognitive strategies, social strategies, and memory strategies. In Yang’s (2007) study, while the most frequently used strategies were also compensation strategies, the ranking of the rest of the categories (social, cognitive, metacognitive, affective and memory strategies in decreasing order of popularity) was different from Lan and Oxford’s (2003) ranking. A few factors may help explain why the ranking of the strategy categories for the present study was different from the above-mentioned studies. First, the present study utilized a strategy questionnaire that was developed specifically for this study. Specifically, it was designed only for writing skills. Those studies with different rankings employed Oxford’s (1990) SILL that only captured general learning strategies. Second, these studies involved different profiles of participants, e.g., primary school pupils (e.g., Lan & Oxford, 2003; Magogwe & Oliver, 2007), and university students (e.g., Baker & Boonkit, 2004; Wharton, 2000; Yang. 2007). Furthermore, culture may also play a role in learners’ preference of strategy use (e.g., Bedell & Oxford, 1996; Gao, 2003). Among all the individual strategies, “paying attention to spellings and grammar when writing” was most frequently used by Singapore upper primary school pupils. This finding is consistent with our observations. Among all the eight subscales/groups, planning strategies (M = 3.37) were

140

Chapter Five

used most frequently by the participants, indicating that the pupils employed more planning strategies as compared with other strategies in their writing process. The least frequently used strategies were selfinitiation strategies (M = 3.05). The findings suggest that pupils might not value self-initiation strategies as highly as other strategies in that selfinitiation strategies are the ones that are utilized by the pupils when they write alone and outside their classroom.

5.2 Relationship between strategy use and English learning outcomes In this section, I will analyze the relationship between strategy use and English learning outcomes. I will report results for each school, respectively. The pattern for the correlation between each variable and language proficiency for all the five schools will be compared. If the pattern is consistent for all the schools, it means that the relationship between strategy use and English learning outcomes is consistent.

5.2.1 Results for School 1 Results of the statistical analyses for School 1 are shown in Table 5.4. The ANOVA analyses showed a significant group effect on five of the eight strategy subscales, i.e., planning, F (2, 626) = 13.82, p = .005; textgenerating, F (2, 626) = 9.93, p = .000; feedback handling, F (2, 626) = 5.28, p = .005; monitoring & evaluating, F (2, 626) = 12.05, p = .000; revising, F (2, 626) = 4.15, p = .016. The significant effect suggests that the five groups of strategies were used differently by the pupils of different English proficiency levels as measured by the z-scores. The post hoc Scheffé tests revealed a general pattern between proficiency and reported strategy use of the five subscales. For revising strategies, the top group reported more strategy use than the middle group, which in turn reported more strategy use than the bottom group (see Table 5.5 for descriptive statistics). However, only the difference between the top and the bottom groups reached statistical significance (top > bottom, p = .021). For the other four subscales, the top group significantly outperformed the middle group (planning, p = .001; text-generating, p = .002; feedback handling, p = .025; monitoring & evaluating, p = .008), whereas the middle group did not outperform the bottom group significantly. The top group also used the four groups of strategies more frequently than the bottom group significantly (planning, p = .000; text-generating, p = .000; feedback handling, p = .017; monitoring & evaluating, p = .000).

Findings and Discussion of Phase One

141

Table 5.4. ANOVA Results of School 1 Source Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error)

SS Self-initiation strategies 1.45 384.297 Planning strategies 12.929 292.766

df

MS

F

2

.725

1.181

.308

626

.614

2

6.464

13.823

.000

626

.468

12.054

.000

9.930

.000

4.148

.016

5.276

.005

1.984

.138

.901

.407

Monitoring & evaluating strategies 11.991 2 5.995 311.361

626

.497

Text-generating strategies 12.277 2

6.138

386.977

626

.618

2

2.559

626

.617

Feedback handling strategies 6.703 2

3.352

Revising strategies 5.118 386.361

397.644 Resourcing strategies 3.003 473.610

626

.635

2

1.501

626

.757

Help-seeking & affect-managing strategies .970 2 .485 336.141

625

.538

p

Chapter Five

142

Table 5.5. Descriptive Statistics of Strategy Use by Proficiency of School 1

Subscale

Bottom group

Middle group

n

n

Self-initiation

206

Planning

206

Monitoring & evaluating

206

Textgenerating Revising

206

Feedback handling Resourcing

206

Help-seeking & affectmanaging

206

206 206

Mean (SD) 2.97 (.80) 3.16 (.71) 3.02 (.72) 3.05 (.80) 3.05 (.80) 3.10 (.80) 2.97 (.90) 3.12 (.71)

207 207 207 207 207 207 207 206

Mean (SD) 3.01 (.81) 3.26 (.67) 3.14 (.70) 3.08 (.78) 3.11 (.81) 3.12 (.81) 2.99 (.85) 3.03 (.77)

Top group n 216

Mean (SD) 3.09 (.74)

216

3.50 (.67)

216

3.35 (.69)

216

3.36 (.78)

216

3.26 (.76)

216

3.32 (.79)

216

3.13 (.86)

216

3.10 (.71)

5.2.2 Results for School 2 Table 5.6 shows results for School 2. A very different pattern between proficiency and strategy use was found. A significant group effect was found only on two subscales, i.e., self-initiation strategies, F (2, 400) = 4.27, p = .015 and text-generating strategies F (2, 399) = 3.99, p = .019 among the three proficiency levels. Surprisingly, the Scheffé tests showed an interesting difference on self-initiation strategies between the groups. The middle group reported using more self-initiation strategies than both the top and the bottom group (see Table 5.7 for descriptive statistics). However, only the difference between the middle and the top groups reached statistical significance (middle > top, p = .019). This finding suggests that in this school the middle group used more self-initiation strategies than the top group. Self-initiation strategies are comprised of such strategies as taking actions to find out how to write effectively, practicing on own accord, analyzing or following good models, and asking for feedback from other people. For text-generating strategies, only the top

Findings and Discussion of Phase One

143

group significantly outperformed the bottom group (top > bottom, p = .026). It should be pointed out although the ANOVA results did not show a statistically significant group effect on planning strategies, F (2, 399) = 2.75, p = .065 and monitoring and evaluating strategies, F (2, 399) = 2.62, p = .074, the group effect on the two subscales approached statistical significance. Table 5.6. ANOVA Results of School 2 Source Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error)

SS Self-initiation strategies 5.213 243.949 Planning strategies 2.285 165.949

df

MS

F

2

2.607

4.274

.015

400

.610

2

1.142

2.747

.065

399

.416

2.620

.074

3.991

.019

1.139

.269

1.006

.367

Monitoring & evaluating strategies 2.202 2 1.101 167.700

399

.420

Text-generating strategies 4.478 2

2.239

223.872

399

.561

2

.752

400

.570

Feedback handling strategies 1.187 2

.593

Revising strategies 1.505 228.144

235.423

399

.590

p

Chapter Five

144

Between Within (error) Between Within (error)

Resourcing strategies 2.537 351.482

2

1.268

400

.879

Help-seeking & affect-managing strategies 1.059 2 .530 210.426

398

1.444

.237

1.002

.368

.529

Table 5.7. Descriptive Statistics of Strategy Use by Proficiency of School 2 Bottom group n Mean (SD) 129 3.12 (.75) 128 3.22 (.56) 128 3.05 (.64)

Middle group n Mean (SD) 138 3.19 (.78) 138 3.38 (.66) 138 3.19 (.64)

Textgenerating Revising

128

138

Feedback handling Resourcing

128

Subscale Self-initiation Planning Monitoring & evaluating

Help-seeking & affectmanaging

129

129 128

2.90 (.70) 3.07 (.73) 3.14 (.77) 3.03 (.90) 3.20 (.72)

138 138 138 138

3.09 (.76) 3.22 (.77) 3.20 (.74) 3.19 (.97) 3.31 (.71)

n

Top group Mean (SD)

136

2.92 (.81)

136

3.37 (.70)

136

3.22 (.66)

136

3.15 (.77)

136

3.16 (.76)

136

3.27 (.79)

136

3.02 (.93)

135

3.20 (.76)

6.2.3 Results for School 3 Results similar to those of School 1 were obtained for School 3. As can been seen in Table 5.8, a group effect was found on six of the eight strategy groups among the three proficiency levels. Similar to School 1, the main effects were significant for planning, F (2, 853) = 20.96, p = .000;

Findings and Discussion of Phase One

145

text-generating, F (2, 853) = 13.85, p = .000; feedback handling, F (2, 853) = 10.54, p = .000; monitoring & evaluating, F (2, 853) = 9.36, p = .000; revising, F (2, 853) = 4.05, p = .018. The different finding between School 1 and School 3 is the significant main effect for resourcing strategies, F (2, 853) = 8.05, p = .000 for School 3. The post hoc Scheffé tests showed that the difference between the top and the middle groups was not statistically significant for the six subscales. However, for planning strategies, significant differences were found between the top and the bottom groups (top > bottom, p = .000) and the middle and the bottom groups (middle > bottom, p = .000) (see Table 5.9 for descriptive statistics). The same pattern was also obtained for text-generating strategies (top > bottom, p = .000; middle > bottom, p = .001), feedback handling (top > bottom, p = .000; middle > bottom, p = .002) and monitoring & evaluating strategies (top > bottom, p = .000; middle > bottom, p = .008). Significant difference was only found between the middle and the bottom groups for revising strategies (middle > bottom, p = .018) and resourcing strategies (middle > bottom, p = .000). Table 5.8. ANOVA Results of School 3 Source Between Within (error) Between Within (error) Between Within (error) Between Within (error)

SS Self-initiation strategies 2.064 547.342 Planning strategies 17.931 364.937

df

MS

F

2

1.032

1.609

.201

853

.642

2

8.966

20.956

.000

853

.428

9.363

.000

13.847

.000

Monitoring & evaluating strategies 8.660 2 4.330 394.487

853

.462

Text-generating strategies 17.082 2

8.541

526.122

853

.617

p

Chapter Five

146

Between Within (error) Between Within (error) Between Within (error) Between Within (error)

Revising strategies 4.927

2

2.464

853

.609

Feedback handling strategies 13.779 2

6.889

519.584

557.418 Resourcing strategies 13.609 721.040

853

.653

2

6.805

853

.845

Help-seeking & affect-managing strategies 1.285 2 .643 436.793

852

4.045

.018

10.542

.000

8.050

.000

1.254

.286

.513

Table 5.9. Descriptive Statistics of Strategy Use by Proficiency of School 3

Subscale

Bottom group

Middle group

n

n

Self-initiation

266

Planning

266

Monitoring & evaluating

266

Textgenerating Revising

266

Feedback handling Resourcing

266

266

266

Mean (SD) 3.05 (.80) 3.22 (.65) 3.06 (.68) 3.00 (.72) 3.10 (.78) 3.06 (.80) 3.02 (.94)

290 290 290 290 290 290 290

Mean (SD) 3.17 (.80) 3.54 (.64) 3.24 (.66) 3.26 (.79) 3.25 (.75) 3.30 (.82) 3.34 (.89)

Top group n 300

Mean (SD) 3.09 (.80) 3.53 (.67)

300 300

3.30 (.70)

300

3.34 (.84)

300

3.20 (.81) 3.36 (.80)

300 300

3.18 (.93)

Findings and Discussion of Phase One

Help-seeking & affectmanaging

265

3.22 (.72)

290

3.27 (.74)

147

300

3.17 (.70)

5.2.4 Results for School 4 Table 5.10 displays the results of statistical analyses for School 4. The general pattern of strategy use for School 4 was similar to that for School 3. Among the eight writing strategy subscales, the same six subscales showed a significant group effect: planning, F (2, 753) = 26.29, p = .000; monitoring and evaluating, F (2, 753) = 6.13, p = .002; text-generating, F (2, 753) = 36.40, p = .000; revising, F (2, 754) = 7.26, p = .001; feedback handling, F (2, 753) = 11.58, p = .000; resourcing, F (2, 753) = 6.06, p = .002. In other words, the six groups of strategies were employed differently by pupils of the different English proficiency levels (see Table 5.11 for descriptive statistics). According to the post hoc Scheffé tests, the top group used planning strategies significantly more frequently than both the middle and the bottom groups (top > middle, p = .000; top > bottom, p = .000). However, the difference between the middle and the bottom groups did not reach a significant level (middle > bottom, p = .060). A similar pattern of strategy use was found for monitoring & evaluating (top > middle, p = .000; top > bottom, p = .000), text-generating (top > middle, p = .000; top > bottom, p = .000), and feedback handling (top > middle, p = .000; top > bottom, p = .000). Significant differences were only found between the top and the bottom groups in their use of revising strategies (top > bottom, p = .001) and resourcing strategies (top > bottom, p = .003). Table 5.10. ANOVA Results of School 4 Source Between Within (error) Between Within (error)

SS Self-initiation strategies 1.543 452.344 Planning strategies 20.690 296.252

df

MS

F

P

2

.772

1.286

.277

754

.600

2

10.345

26.294

.000

753

.393

Chapter Five

148

Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error)

Monitoring & evaluating strategies 5.054 2 2.527 310.329

753

.412

Text-generating strategies 39.306 2

19.653

406.561

753

.617

2

4.120

754

.568

Feedback handling strategies 14.168 2

7.084

Revising strategies 8.240 427.999

460.610 Resourcing strategies 9.760 607.178

753

.612

2

4.880

753

.805

Help-seeking & affect-managing strategies 1.547 2 .773 360.820

753

6.132

.002

36.400

.000

7.258

.001

11.581

.000

6.060

.002

1.614

.200

.479

Table 5.11. Descriptive Statistics of Strategy Use by Proficiency of School 4

Subscale

Bottom group

Middle group

n

n

Self-initiation

258

Planning

258

Monitoring & evaluating

258

Mean (SD) 3.02 (.81) 3.18 (.65) 3.02 (.69)

273 272 272

Mean (SD) 2.96 (.78) 3.31 (.64) 3.08 (.65)

Top group n 226 226 226

Mean (SD) 3.07 (.73) 3.59 (.59) 3.22 (.58)

Findings and Discussion of Phase One

Textgenerating Revising

258

Feedback handling Resourcing

258

Help-seeking & affectmanaging

258

258 258

2.87 (.70) 3.00 (.77) 3.03 (.75) 2.97 (.85) 3.11 (.65)

272 273 272 273 273

2.91 (.73) 3.10 (.73) 3.05 (.81) 3.08 (.92) 3.11 (.73)

149

226 226 226 226 225

3.39 (.77) 3.26 (.76) 3.34 (.78) 3.25 (.93) 3.21 (.69)

5.2.5 Results for School 5 The ANOVA results of School 5 can be found in Table 5.12. Four subscales of strategies showed a significant group effect, indicating that pupils from different proficiency groups generally differed in their use of the four groups of strategies, i.e., planning, F (2, 522) = 9.96, p = .000; text-generating, F (2, 522) = 23.18, p = .000; revising, F (2, 522) = 4.25, p = .015; feedback handling, F (2, 522) = 7.97, p = .000. For planning strategies, the post hoc analysis by Scheffé tests showed a significant difference between the top and the bottom groups (top > bottom, p = .000) as well as between the middle and the bottom groups (middle > bottom, p = .026) (see Table 5.13 for descriptive statistics). With regard to textgenerating strategies, the top group reported more frequent use than the middle group (top > middle, p = .047), which in turn significantly outperformed the bottom group (middle > bottom, p = .001). For feedback handling strategies, the top group reported more frequent strategy use as compared to both the middle group (top > middle, p = .025) and the bottom group (top > bottom, p = .001). Difference in the use of revising strategies was only found between the top and the bottom groups (top > bottom, p = .017).

Chapter Five

150

Table 5.12. ANOVA Results of School 5 Source Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error) Between Within (error)

SS Self-initiation strategies .875 348.306 Planning strategies 8.515 223.181

df

MS

F

2

.437

.657

.519

523

.666

2

4.257

9.958

.000

522

.428

3.177

.043

23.183

.000

4.245

.015

7.968

.000

1.543

.215

.437

.646

Monitoring & evaluating strategies 2.845 2 1.422 233.728

522

.448

Text-generating strategies 26.192 2

13.096

294.881

522

.565

2

2.379

522

.560

Feedback handling strategies 8.952 2

4.476

Revising strategies 4.757 292.477

293.255 Resourcing strategies 2.287 387.049

522

.562

2

1.144

522

.741

Help-seeking & affect-managing strategies .493 2 .246 293.504

521

.563

p

Findings and Discussion of Phase One

151

Table 5.13. Descriptive Statistics of Strategy Use by Proficiency of School 5

Subscale

Bottom group

Middle group

n

n

Self-initiation

197

Planning

196

Monitoring & evaluating

196

Textgenerating Revising

196

Feedback handling Resourcing

196

Help-seeking & affectmanaging

196

196 196

Mean (SD) 3.08 (.82) 3.26 (.66) 3.09 (.74) 2.87 (.78) 3.10 (.76) 3.14 (.76) 3.09 (.86) 3.17 (.74)

148 148 148 148 148 148 148 148

Mean (SD) 3.00 (.78) 3.45 (.70) 3.21 (.62) 3.19 (.76) 3.17 (.71) 3.21 (.77) 3.22 (.90) 3.10 (.79)

Top group n

Mean (SD)

181

2.99 (.84)

181

3.56 (.61)

181

3.26 (.63)

181

3.40 (.71)

181

3.32 (.76)

181

3.44 (.72)

181

3.23 (.83)

180

3.12 (.73)

5.3 Patterns of reported strategy use across the participating schools The findings from the ANOVA analyses conducted for the five participating schools suggest that overall there was systematic difference among the pupils of the different proficiency levels grouped according to their English language results. Planning strategies were found to be consistently significant in four of the five schools. In particular, the top group used planning strategies more often than the bottom group in all the four schools. In two of the four schools, the top group was found to outperform the middle group in their use of planning strategies, whereas in the other two, the middle group planned more often than the bottom group (middle > bottom). With regard to the monitoring and evaluating strategies, a significant relationship between strategy use and English proficiency was found in

152

Chapter Five

three of the five participating schools, with the top group outperforming the bottom group in all the three, the top outperforming the middle group in two of the three schools, and the middle outperforming the bottom group in one school. For the text-generating strategies, a significant positive relationship was found between strategy use and English proficiency in all the five schools. All the five schools exhibited a “top > bottom” pattern, indicating that the pupils from the top group employed more text-generating strategies than their counterparts from the bottom group. Three of the five schools showed a “top > middle” pattern, while two schools displayed a “middle > bottom” pattern. Reported use of the revising strategies exhibited a significant positive relationship with English proficiency in four schools. However, the pattern was less consistent. A pattern of “top > bottom” was found only in two schools, a pattern of “middle > bottom” in two schools, and a pattern of “top > middle” in one school. The feedback handling strategies were found to be positively correlated with English proficiency in four schools. The top group exhibited their superiority in strategy use compared to the bottom group in the four schools. It was also found that the top group outperformed the middle group in three of the four schools and the middle group reported more frequent use than the bottom group in one school. The resourcing strategies showed a significant group effect only in two schools, with a “middle > bottom” in one school and a “top > bottom” pattern in the other. As for the self-initiation strategies, difference in strategy use was only found in one school with the middle group outperforming the bottom group. No difference was found among the three proficiency levels in their use of help-seeking and affective strategies in all the five participating schools, suggesting that this subscale was not a good predictor of English proficiency for Singapore primary school pupils. As was discussed above, the planning strategies, feedback handling strategies and revising strategies were consistently found to be positively correlated with English proficiency in four of the five participating schools (80% of the participants), and text-generating strategies in all the five schools (100%). These four subscales were significant positive predictors of English language learning outcomes in the four schools.

Findings and Discussion of Phase One

153

It is important to highlight that these findings corroborate the literature reviewed in chapter 2. To be specific, planning is a crucial process for effective writing; one of the three primary cognitive processes in Flower and Hayes’s (1981) writing model is planning. Empirical studies also showed effective writers’ superiority in their use of planning strategies (e.g., Chamot & El-Dinary, 1999; Chien, 2012; Gu et al., 2005; O’Malley & Chamot, 1990; O’Malley et al., 1985a; Vandergrift, 2003a; Victori, 1999). The feedback handling strategies could also predicate well the English proficiency of Singapore primary school pupils. As was reviewed in chapter 2, peer feedback does not only help the writers who receive feedback on their writing but also help those who provide feedback in the shared process among learners who read, critique, and provide feedback on each other’s work. Empirical research (e.g., Hu, 2005, Min, 2006) also showed many benefits accorded to by peer review. The text-generating strategies were also found to be an important process associated with better writing competence (e.g., Chien, 2012; Glaser & Brunstein, 2009; McCutchen, 2006). Numerous research studies identified differences between effective strategies and their ineffective counterparts in their use of revising strategies (e.g., Chien, 2012; Graham, 2006; MacArthur, 2007; Victori, 1999). In addition, revising is one of the three primary cognitive processes (i.e., planning, translating and reviewing [revising and editing]) in Flower and Hayes’s (1981) cognitive model of writing. The research evidence of this phase was able to establish the target strategies that should be taught to Singapore primary school pupils in Phase 2, which included planning strategies, feedback handling strategies, text-generating strategies and revising strategies. The non-target strategies were the other subscales/groups of strategies on the questionnaire, i.e., self-initiation strategies, monitoring and evaluating strategies, resourcing strategies, and social/affective strategies.

5.4 Conclusion The results of Phase 1 identified consistent relationships between strategy use and English proficiency on four groups of writing strategies, i.e., planning strategies, feedback handling strategies, text-generating strategies, and revising strategies. It was made clear previously if the patterns of

154

Chapter Five

strategy use were found to be consistent in the majority of the schools, there was a good reason to believe that these strategies are good predictors of language learning outcomes. Therefore, the findings can serve as one source of input for deciding on the intervention strategies.

CHAPTER SIX FINDINGS AND DISCUSSION OF PHASE TWO

This chapter will report and discuss the findings from Phase 2, which investigated the effectiveness of strategy-based writing instruction in Singapore primary schools. As was made clear previously, the design of Phase 2 was partly informed by the results obtained from Phase 1. Phase 2 aimed to address the following two research questions: 1. In what ways does strategy-based writing instruction help pupils improve their writing competence in English? 2. In what ways does such instruction help improve pupils’ use of the target writing strategies?

6.1 The quantitative results Table 6.1 presents the descriptive statistics comparing the experimental and the control groups on the pre-test, the post-test and the delayed test. It was already pointed out in chapter 4 that a quasi-experimental design was adopted for the present phase. There might be differences between the experimental and the control groups on the covariate, i.e., the pre-test score, since intact classes were assigned to both the experimental and control groups. Therefore, an analysis of covariance (ANCOVA) with repeated measures was adopted with the pre-test score as the covariate in order to remove the pre-existing differences between the experimental and the control groups if any (see Dörnyei, 2007; Huck, 2008). Descriptive statistics of the three tests were also obtained. Means, standard deviations, F rations, and correlations, will be reported accordingly. The effect size or strength of relationship between the covariate (i.e., independent variable) and the dependent variable will also be reported as an important measure of the effects of the treatment according to the APA requirements (see Nesbit & Hadwin, 2006).

Chapter Six

156

Table 6.1. Descriptive Statistics of Pre-, Post-, and Delayed Tests Treatment Experimental

Control

Pre-test Mean SD 47.998 7.438 (n = 221) 47.482 6.045 (n = 221)

Post-test Mean SD 52.106 6.479 (n = 221) 48.106 5.579 (n = 221)

Delayed test Mean SD 50.889 5.911 (n = 190) 48.490 6.113 (n = 201)

6.1.1 Effects of the SBI on writing competence As was noted by Huck (2008), the covariate variable must be related to the dependent variable for an ANCOVA. Therefore, an ANCOVA was run to confirm whether the pre-test was conceptually related to the post-test and the delayed test. The pre-test (covariate) scores were significantly related to both the post-test and the delayed test scores, F (1, 389) = 194.189, r = .51, p < .001 for the post-test, r = .50, p < .001 for the delayed test. After the initial differences (i.e., pre-test scores) were controlled for, the experimental and the control groups were compared on the post- and the delayed test results. As shown in Table 6.1, there were significant differences between the experimental and the control groups, F (1, 439) = 58.668, p < .001, Ș2 = .118 for the post-test, and (1, 388) = 21.465, p < .001, Ș2 = .052 for the delayed test. Dörnyei (2007, p. 221) and (Huck 2008, p. 246) proposed three levels for the effect size of an ANCOVA: small effect = 0.01, moderate effect = 0.06, and large effect = 0.14. The results showed that the experimental group (M = 52.106) did significantly better than the control group (M = 48.106) on the post-test, with an effect size (0.118) close to a large effect (0.14). On the delayed test, although the pattern of superior performance by the experimental group was sustained over one month’s time, the difference between the experimental group (M = 50.889) and control groups (M = 48.492) became smaller, with the effect size decreasing to .051, close to a moderate effect size (0.06). Such results indicated that the effect of the SBI was still sustained one month after the intervention.

Findings and Discussion of Phase Two

157

6.1.2 Effects of the SBI on use of writing strategies As was mentioned in chapter 4, the writing strategy questionnaire included eight variables (subscales): self-initiation, planning, monitoring and evaluating, feedback handling, revising, text-generating, resourcing, and social/affective strategies. In this section, I will report the results with regards to the effects of the SBI on the participants’ strategy use. The independent-samples t tests (see Table 6.2) showed that there was no significant difference on the reported strategy use between the experimental and the control groups before the intervention, showing that the two groups were comparable. Table 6.2. T Test on Frequency of Strategy Use Comparing Experimental and Control Groups Strategy

Self-initiation Planning Monitoring & evaluating Text-generating Revising Feedback handling Resourcing Help-seeking & affectmanaging

Mean n (SD) Experimental group 2.71 221 (.74) 3.32 220 (.65) 3.13 220 (.68) 3.06 220 (.76) 3.07 221 (.74) 3.11 219 (.82) 2.99 221 (.94) 3.00 221 (.69)

Mean n (SD) Control group

t test t

p

2.84 (.76)

218

1.75

.082

3.36 (.64)

217

.66

.513

3.17 (.69)

219

.66

.512

3.04 (.75)

219

.771

3.13 (.72)

218

.291 .858

.392

3.10 (.77)

219

-.16

.873

3.06 (.91)

218

.78

.434

3.04 (.76)

219

.62

.536

It was also made clear in chapter 4 that planning, feedback handling, revising, and generating text were the target strategies (variables), while self-initiation, monitoring and evaluating, resourcing, and social/affective were the non-target strategies. If the intervention is targeted at specific

Chapter Six

158

variables (strategies), the intervention should not yield any effect on those non-target strategies. In other words, the SBI should only bring about positive effects to the target strategies, instead of the non-target ones. For each group of the strategies on the questionnaire, I carried out both descriptive statistics and an ANCOVA test. For the sake of clarity, I will present results for the non-target strategies first and then the target strategies. Self-initiation strategies (non-target) The ANCOVA showed that the experimental and the control groups did not differ significantly in their use of self-initiation strategies after the intervention, F (1, 426) = .130, p = .719, Ș2 = .000. The covariate was significantly related to post-intervention strategy use, F (1, 426) = 134.356, r = .49, p < .001. However, it is interesting to note that the adjusted mean of self-initiation strategies by the control group dropped from M = 2.838 (pre-intervention score) to M = 2.807 (post-intervention score), whereas it increased for the experimental group from M = 2.713 to M = 2.872 (see Table 6.3). A paired-samples t test was conducted to compare the means of the experimental group’s strategy use after the intervention. The result suggests there was no significant increase in the experimental group’s use of self-initiation strategies after the intervention, t = 1.079, p = .282. Table 6.3. Descriptive Statistics of Self-initiation Strategies Pre-intervention Group

Post-intervention

n Mean

SD

Mean

SD

Experimental

217

2.713

.744

2.782

.789

Control

212

2.838

.762

2.807

.781

The use of initiation strategies fell into the range of medium use (Oxford, 1990). In contrast, Phase 1 also showed that the participants employed this group of strategy at a range of medium frequency.

Findings and Discussion of Phase Two

159

Monitoring and evaluating strategies (non-target) The monitoring and evaluating strategies included such strategies as maintaining task focus as planned, reading aloud to look for mistakes, attending to specific aspects of writing, e.g., spellings and grammar, and comparing one’s own writing with others’. The descriptive statistics (see Table 6.4) show that both the experimental and the control groups scored higher on the post-test. The experimental group’s adjusted mean increased from M = 3.125 to M = 3.253, and the control group from M = 3.168 to M = 3.246. Both groups’ strategy use fell into a medium range as shown on the post-intervention questionnaire. Table 6.4. Descriptive Statistics of Monitoring and Evaluating Strategies Pre-intervention

Post-intervention

Group

n

Experimental

217

Mean 3.125

SD .679

Mean 3.253

SD .674

Control

211

3.168

.689

3.246

.668

The ANCOVA indicated that the covariate was significantly correlated with the participants’ use of monitoring and evaluating strategies on the post-intervention questionnaire, F (1, 425) = 131.220, p < .001, r = .485. As was expected, no significant treatment effect was found, F (1, 425) = .281, p = .596, Ș2 = .001. The paired-samples t test suggested that there was no significant increase in the experimental group’s use of monitoring and evaluating strategies after the intervention, t = 2.63, p = .09. Resourcing strategies (non-target) As was expected, no positive treatment effect on the use of resourcing strategies was found after the intervention. As shown in Table 6.5, the experimental group showed minimal increase on their use of resourcing strategies from M = 2.986 to M = 2.996. For the control group, the mean score decreased from M = 3.056 to M = 3.039. Overall, the use of this group of strategies fell into the medium frequency.

Chapter Six

160

Table 6.5. Descriptive Statistics of Resourcing Strategies Pre-intervention

Post-intervention

Group

n

Experimental

216

Mean 2.986

SD .942

Mean 2.996

SD .909

Control

211

3.056

.913

3.039

.918

Although the ANCOVA revealed that the covariate (use of resourcing strategies on the pre-intervention questionnaire) was significantly related to the use of resourcing strategies on the post-intervention questionnaire, F (1, 424) = 124.686, p < .001, r = .477, there was no statistically significant treatment effect, F (1, 424) = .143, p = .706, Ș2 = .000. The paired-samples t test suggested that the experimental group’s use of this group of strategies did not change significantly, t = 1.03, p = .918. Therefore, the findings suggested that the intervention did not bring about significant change in strategy use for the experimental group. Social/Affective strategies (non-target) The final group of non-target strategies on the questionnaire was social/affective strategies. The descriptive statistics (see Table 6.6) show very marginal improvement on their strategy use for both groups. The experimental group’s use of social/affective strategies increased from M = 3.001 to M = 3.033, and the control group from M = 3.044 to M = 3.097, with both groups falling into the range of medium frequency. Table 6.6. Descriptive Statistics of Social/Affective Strategies Pre-intervention

Post-intervention

Group

n

Experimental

217

Mean 3.001

SD .692

Mean 3.033

SD .757

Control

210

3.044

.755

3.097

.698

There was no statistically significant difference on the use of social/affective strategies between the experimental group and the control groups after the intervention. The covariate (use of resourcing strategies

Findings and Discussion of Phase Two

161

on the pre-intervention questionnaire) was significantly related to the participants’ post-intervention strategy use, F (1,424) = 110.830, p < .001, r = .456. The treatment effect was found to be statistically insignificant, F (1,424) = 423, p = .516, Ș2 = .001. In addition, no significant change in the experimental group’s use of social/affective was found, t = .512, p = .609 after the intervention. Planning strategies (target) The first group of target strategies on the questionnaire was planning strategies. The descriptive statistics revealed that both the experimental group and the control group used their planning strategies more frequently after the intervention. As shown in Table 6.7, the adjusted mean of strategy use increased from M = 3.352 to M = 3.516 for the experimental group and from M = 3.389 to M = 3.446 for the control group. The covariate (mean score of planning strategies) on the pre-intervention questionnaire was significantly related to the participants’ use of planning strategies on the post-intervention questionnaire, F (1, 424) = 110. 134, p < .001, r = .45. Table 6.7. Descriptive Statistics of Planning Strategies Pre-intervention

Post-intervention

Group

n

Experimental

217

Mean 3.353

SD .812

Mean 3.516

SD .820

Control

210

3.389

.768

3.446

.771

However, the effect of treatment for the experimental group was found to be insignificant, F (1, 424) = 2. 603, p = .107, Ș2 = .006. Therefore, the analysis suggested that the SBI did not yield a significant treatment effect on the experimental group’s use of planning strategies. Feedback handling strategies (target) The second group of target strategies was feedback handling strategies, which were comprised of “trying to remember or write other people’s suggestions for future use”, “thinking carefully about other people’s suggestions for own compositions” and so on. Table 6.8 shows that the mean score of feedback handling strategy use for the experimental group

Chapter Six

162

grew from M = 3.104 to M = 3.245. The score for the control group decreased from M = 3.089 to M = 3.079. Table 6.8. Descriptive Statistics of Feedback Handling Strategies Pre-intervention

Post-intervention

216

Mean 3.014

SD .826

Mean 3.245

SD .814

208

3.089

.741

3.079

.768

Group

n

Experimental Control

The ANCOVA revealed that the intervention on feedback handling strategies was successful. The covariate was significantly related to the participants’ post-intervention use of feedback handling strategies, F (1, 421) = 67.716, p < .001, r = .37. There was a positive treatment effect, F (1, 421) = 4.991, p = .026, Ș2 = .019. The analysis suggested that the participants in the experimental group used employed more feedback handing strategies than their counterparts in the control group as a result of the SBI. Revising strategies (target) Both the experimental and the control groups made good progress in their use of revising strategies, with the experimental group increasing from M = 3.084 to M = 3.275 and the control group from M = 3.132 to M = 3.168 (see Table 6.9). Table 6.9. Descriptive Statistics of Revising Strategies Pre-intervention

Post-intervention

Group

n

Experimental

217

Mean 3.084

SD .735

Mean 3.275

SD .728

Control

212

3.132

.721

3.168

.706

There was a statistically significant treatment effect for revising strategies, as shown by the ANCOVA results. The covariate (use of revising strategies on the pre-intervention questionnaire) was found significantly related to the post-intervention revising strategies use, F (1, 426) =

Findings and Discussion of Phase Two

163

119.924, p < .001, r = .46. The pupils in the experimental group used more revising strategies as a result of the SBI, F (1, 426) = 84.346, p = .038, Ș2 = .010, with a small effect size (Dörnyei 2007, p. 221; Huck 2008, p. 246). Text-generating strategies (target) The same pattern of positive treatment effect (as that of revising strategies) was also found for text-generating strategies. The experimental group’s use of text-generating strategies grew from M = 3.024 to M = 3.336, and the control group’s use from M = 3.059 to M = 3.139 (see Table 6.10). Table 6.10. Descriptive Statistics of Text-generating Strategies Pre-intervention

Post-intervention

n

Group

Mean

SD

Mean

SD

Experimental

218

3.024

.722

3.336

.751

Control

211

3.059

.757

3.139

.738

The ANCOVA showed evidence of a positive treatment effect, F (1, 426) = 8.074, p = .005, Ș2 = .012. The covariate was significantly related to the participants’ use of text-generating strategies on the post-intervention questionnaire, F (1, 426) = 120.167, p < .001, r = .47. As such, the analysis revealed that the participants in the experimental group employed text-generating more frequently than their counterparts in the control group after the intervention.

6.1.3 Discussion of the quantitative results The project investigated the effects of strategy-based writing instruction for young writers on their strategy use as well as their English writing competence in three Singapore primary schools. Using a quasiexperimental design, the present study found that the intervention led to significant improvements in the participants’ writing competence and the target writing strategies. Similar positive treatment effects of SBI were also reported by De Silva (2015).

164

Chapter Six

The findings are important for a number of reasons. First, there was an overall impact of the strategy-based writing instruction on the participants in the experimental group in their use of the target writing strategies right after the intervention. In addition, the experimental group made statistically significant gains on the post-test, which means the SBI had a direct treatment effect on the participants’ writing competence (research question 1). The good performance of the experimental group still remained as shown on the delayed test mean score one month after the intervention ended. As was already pointed out, previous research (see Graham, 2006b for details) has shown that strategy-based writing instruction led to positive treatment effects on the quality of the participants’ writing. The present study lends further support to such findings. Graham (2006b) found in his meta-analysis that the treatment effect was maintained with a large effect size in addition to the large effect size as measured immediately after the intervention in seven of the 20 studies. Similar to the findings in Graham (2006b), the effect size of the present study became smaller (at the delayed test) four weeks after the intervention. Fidalgo, Torrance, and García (2008) also reported similar findings. Second, there was also statistically significant improvement by the participants in the experimental group as compared to those in the control group in their use of three of the four target groups of strategies, i.e., feedback handling strategies, revising strategies, and text-generating strategies (research question 2). The treatment effect (i.e., .010 to .019) fell into the range of a small effect size (Dörnyei 2007; Huck 2008). As expected, there was significant treatment effect for feedback handling strategies, revising strategies, and text-generating strategies. However, no difference was found in the use of planning strategies. A small effect size is expected as only a subset of each type of writing strategies on the questionnaire was taught. In fact, only about 50% of the writing strategies in each group of writing strategies were taught in the SBI. Therefore, although the SBI saw improvements in the participants’ use of the writing strategies, the improvement could not raise the overall score of each subscale in the questionnaire. In addition, as was discussed previously, orchestration characterizes successful language learning, thus writing competence (e.g., Anderson, 2005; Zhang et al., 2008). However, the writing strategy questionnaire elicited only frequency data on the participants’ strategy use and could not capture the quality of their strategy use, i.e., orchestration. Therefore, it is acceptable to find no significant difference in the use of planning strategies between the experimental and

Findings and Discussion of Phase Two

165

the control groups. When I turn to the qualitative analyses of the student interviews in section 6.2, it may well be true that the pupils plan their writing with a better quality. Third, the inclusion of non-target variables was to test whether mere participation in a program rather than the intervention itself accounts for the positive intervention effects if any. Indeed, no intervention effects were found on the non-target variables, i.e., self-initiation strategies, F (1, 426) = .130, p = .719, Ș2 = .000, monitoring and evaluating strategies, F (1, 425) = .281, p = .596, Ș2 = .001, resourcing strategies, F (1, 424) = .143, p = .706, Ș2 = .000, and social/affective strategies, F (1,424) = 423, p = .516, Ș2 = .001. In addition, the paired-samples t tests suggested that there were no gains on the non-target variables between the pre-intervention questionnaire survey and the post-intervention questionnaire survey for the experimental group. Such findings lend further support to the validity of the intervention. In sum, the quantitative results show robust evidence that the SBI was successful in helping the pupils to improve their writing competence as well as their strategy use. The traditional approaches to teaching writing may remain at a disadvantage as the findings suggested that there was still a treatment effect one month after the intervention. The pupils in the experimental group employed more target strategies, i.e., planning strategies, feedback handling strategies, text-generating strategies and revising strategies, whereas no significant change in strategy use was found in the pupils in the control group.

6.2 The qualitative results The strategy-based writing instruction brought about change in the participants’ strategy use not only quantitatively but also qualitatively. As was already mentioned in 4.2.3, teacher and pupil interviews were conducted to investigate qualitative effects of the SBI on the pupils’ strategy use. This is to provide further evidence as well as the relationship between strategy use and writing proficiency (Fidalgo et al., 2008). Only a few cases of the student interviews will be reported here as it is not possible to report all the qualitative data.

Chapter Six

166

6.2.1 Student interviews In this section, I will only focus on the pupils’ reported strategy use. It was found that the SBI not only helped the pupils increase their repertoire of writing strategies but also helped develop awareness of the usefulness of strategies they had not been using before the SBI. In general, the pupils became more aware that they could employ the writing strategies they were already using before the intervention more effectively. For example, during the pre-intervention interview, Felicia (high proficiency) mentioned that effective writers always put their heart into writing (see Excerpt 1) when she was asked about “why do you think that some students can write good compositions but others can’t…?” Ineffective writers “don’t really take planning their composition very hard.” It seems that Felicia understands the importance of planning, but to a limited extent. In addition, she only mentioned that effective writers normally do planning. Excerpt 1 I: … why do you think that some students can write good compositions but other...other students they don't write good compositions? P: Mm because some they don't put their heart into writing it. They just anyhow do it. I: What do you mean by “put their heart”? P: Like they really know what they are writing and they really plan it very well. ……………………………………………………………………………… I: Ok. And the other students who are not writing well, they…? P: They don't really take planning their composition very hard. I: Ok, good good. What else do you think? P: Nothing. -Felicia’s pre-intervention interview, high proficiency girl

In the post-intervention interview (see Excerpt 2), however, she was able to talk about planning strategies more extensively. For example, she mentioned that good writers remember their strategies and apply them in their writing. It is also interesting to note that she used the term “strategy” in her interview.

Findings and Discussion of Phase Two

167

Excerpt 2 I: … Ok. … but others cannot? P: Uh because they do their planning well and they know what they are writing. I: Why do you say that they do their planning well? What do you mean? P: They remember their strategies and they plan it and do it into their compositions. I: Ok so what do you mean by strategies? P: Uh the setting of goals. I: Good. P: characters. I: Right. P: the settings and event. -Felicia’s post-intervention interview, high proficiency girl

After the intervention, Felicia also mentioned that good writing involved “checking of grammar mistakes”, a revising strategy that was not mentioned before (see Excerpt 3). She became more aware of the importance of the organization of a composition as well. Excerpt 3 I: Good, what else can you think of? P: Mm checking of grammar mistakes. I: Good good. P: And organize your composition. -Felicia’s post-intervention interview, high proficiency girl

Felicia was also aware that planning involved such strategies as setting of goals, the setting and event, characters and so on (see Excerpt 2). Setting the goals is the most useful writing strategy for her as “that is the main thing in the composition…in writing.” (see Excerpt 4) Excerpt 4 I: …. Uh so uh which of those strategies do you think are very useful to you personally? P: Mm about setting the goals. I: Why? P: Because that is the main thing in the composition...about what is your goal in writing. I: Right right. P: And you will know what to write. -Felicia’s post-intervention interview, high proficiency girl

Chapter Six

168

When asked to explain what goal setting meant (see Excerpt 5), Felicia mentioned that she learned about goal setting in the intervention as she mentioned goal setting was about “Uh… about who, what, why, when, and how”. Excerpt 5 I: uh huh. And what is your writing goal for...I mean for for you? I don't understand. I want you to teach me how to set goals. How please? P: Uh… about who, what, why, when, and how. I: Right right. Ok so later you are going to show me about this ok. Later. I mean after some questions ok? So do you think the use of writing strategies can improve your writing? P: Ya I: Why? P: Because planning helps you to uh write your composition properly, organize your composition and you know what you are going to write in your composition after you plan. -Felicia’s post-intervention interview, high proficiency girl

After the SBI, Felicia showed a larger repertoire of writing strategies and was able to discuss what the strategies were and how the strategies could be used successfully. As compared with her pre-intervention interview, Felicia not only understood the importance of planning, but also understood the conditions on how planning could be done successfully. Another pupil (low proficiency), Angelina, kept silent when asked whether she had heard of writing strategies or good methods for improving writing before the SBI (see Excerpt 6). Like many ineffective student writers, the only strategy that she thought that could help improve her writing was reading English books or speaking English at home. Excerpt 6 I: ……. have you heard of writing strategies before? Any good ways, any good writing ways you know, writing method before? P: (Silent)……. I: Never, ok…why do you think some students can write very good compositions but some students are slow in writing compositions? They are not so good in writing? What are the reasons? P: They maybe read so much books or talk in English at home and maybe talk English with the friends also. -Angelina’s pre-intervention interview, low proficiency girl

Findings and Discussion of Phase Two

169

In comparison, the post-intervention interview revealed that she became aware of a few writing strategies, e.g., overall goals, specific goals and organization. She also mentioned settings (see Excerpt 7). Excerpt 7 I: … So, have you heard of any writing strategies? Writing strategies? P: Yes, quite a bit lah. I: … Emm… can you tell me the writing strategies that you know. P: Mmm… like overall goals, overall… specific goals, organization, emm… what’s that huh? What’s that? Ah… I: Right, right, right. P: Then emm… settings…. -Angelina’s post-intervention interview, low proficiency girl

The excerpt shows that Angelina reported applying some writing strategies (i.e., setting overall and specific goals) that she learned. She also understood importance of goal setting (see Excerpt 8), as she mentioned “…if I never use with goals…, maybe I will write out of point.” Excerpt 8 I: Ok never mind. So, which of those things have you used recently? P: Mmm… overall goals and specific goals. Also, I used it to my compositions. I: Every time. P: Sometimes lah. I: Yah. So which are the things that you find very useful to your writing? P: Mmm… I think like overall goals and specific goals is… specific goal is fine. I: Why? P: Yah, because if I never use with goals, then I don’t know…. Then when I write, maybe I will write out of point. -Angelina’s post-intervention interview, low proficiency girl

Fei Hong is another case to show the effects of the SBI. Prior to the intervention, the most important things that she would pay attention to were grammar and spelling (see Excerpt 9). She said “…make sure I don’t have any grammar mistakes.” Reading more storybooks was a useful way as well. Her strategies were only limited to paying attention to grammar and spelling and reading story books.

Chapter Six

170 Excerpt 9

I: Oh good good. That's good. What are the most important things you pay attention to while you are writing? P: Mm I make sure I don't make any grammar mistakes. I: Mm… hm… P: And...and I spell them correctly. I: Mn… hm…. What else? The most important thing that you pay attention to. P: Nothing. I: Nothing? Ok. Have you got any special ways to improve your writing? P: Just read more storybooks. -Fei Hong’s pre-intervention interview, high proficiency girl

The post-intervention interview showed that she reported more strategy use (see Excerpt 10). She would think/plan what kind of story she would write and how she would start it. In addition, she would check whether she followed her plan or not besides checking grammar and spelling. Excerpt 10 P: Uh I will plan what I will...what kind of story I will write and how I start it. Who are the characters, the settings and the problems that the characters are facing and what is the ending? And then I start writing and make sure I do not make any grammar mistakes and when I finish, I check through thoroughly. -Fei Hong’s post-intervention interview, high proficiency girl

Jia Min, a low proficiency girl, could not describe what she would actually do for writing a composition before the SBI. Excerpt 11 shows that she would try to relax by telling herself that the composition was easy. But when the composition was really difficult, she would be very careful in checking for mistakes in her writing. Excerpt 11 I: Before you write what do you think? P: I will tell myself, this is very easy then it will be easy. I: Good, good…what else? P: If is really difficult, I will…(pause)…think hardly and see whether it is correct. I: Huh…huh… P: If wrong, I will correct my mistakes.

Findings and Discussion of Phase Two

171

I: Huh…huh…this is before you write, right? P: Hmm… -Jia Min’s pre-intervention interview, low proficiency girl

The analysis of her post-intervention interview (see Excerpt 12) shows that Jia Min demonstrated many more strategic behaviors as compared to those in her pre-intervention interview. She would plan the content before writing, which included assessing the audience and choosing the right words. Excerpt 12 I: … So what are the other things that you will think or do before you start to write the composition on the copy that I just showed you? P: To plan the content. I: Ah… what do you mean by plan the content? P: Depend on what I have to write err… assess the audience, the right words. I: …… -Jia Min’s post-intervention interview, low proficiency girl

In terms of the characters, she would think of their expressions and what they are doing (see Excerpt 13). She also talked about how and where the events took place in the story. Therefore, the analysis shows she was able to describe how she would employ the strategies when writing as a result of the intervention. Excerpt 13 P: Err…see how the people… the expression. I: Mmm. P: And see what they are doing. I: Now actually do you plan right, ok… why do you think about all these? P: Mmm…… I: So you were saying that you plan your content right, … and you said…you will actually think about who are in your story. These people are called characters, right? So what else do you think about? P: Events. I: Events... ok. So what do you mean by events here? P: Err… how did the story take place. -Jia Min’s post-intervention interview, low proficiency girl

Besides the above-mentioned findings, another finding that emerged from the analysis is the pupils’ reported use of feedback obtaining and handling

Chapter Six

172

strategies. I will now refer to the analyses of some of the transcripts in which the pupils reported how they obtained and handled feedback from their teachers, family members or friends. As was already made clear in section 3.2, complimentary to the cognitive view of LLS is the sociocultural view, which involves the social mediations between the learners, the teachers or their family members. Luke, a high proficiency boy, talked at quite a few points about the involvement of his teacher, friends and family in his writing process (see Excerpt 14). A key strategy for Luke was asking them for help when he had a problem. He also mentioned in the interview he would ask the teacher, friends or family members when he encountered a problem in writing. Excerpt 14 I: Ok, how do you deal with the difficulties …? P: We asked teacher or your friend or your family. -Luke’s post-intervention interview, high proficiency boy

For him, the people resources around him also appeared quite important when he finished writing a composition as later in the interview he mentioned he would ask his friends or teacher for suggestions on how his composition could be improved (see Excerpt 15). He also employed another strategy, checking the suggestions against his writing goals. He did not just follow their suggestions blindly. If the suggestions did not align with his specific goals, he would not follow those suggestions. Excerpt 15 I: …will you show your composition to someone else for suggestions? P: My friends. I: …you will only show your composition to your friends? P: …and sometimes to my teacher. I: Teacher…ok, do they give you suggestions about how to make your writing better? P: Yes. I: Do you follow their suggestions? P: Yes sometimes. I: Sometimes…why sometimes you don’t follow their suggestion? P: Because sometimes their suggestions are wrong. I: …How do you know that their suggestions are wrong? P: They do not go according to the specific goals. -Luke’s post-intervention interview, high proficiency boy

Findings and Discussion of Phase Two

173

Likewise, Ben, an intermediate proficiency boy also reported using feedback obtaining and handling strategies (see Excerpt 16). He would get his brother to check his composition as he understood that his brother would help him write better compositions in the future. He also revealed that he would not always follow their suggestions as he explained “…, sometimes I modify” when he found their suggestions were not useful to him. Excerpt 16 I: Will you show your composition to anyone? P: Yah, I let them check, see whether… I: Who…who do you think will show it to? P: Friends and family. I: Family, your parents, your sister and your brother? P: Brother. I: Your brother, ok…do you think you can get some suggestion from them? P: Yah, so that I will have like better compositions next time I write about the same topic. I: Ok…do you mean you will follow all their suggestions, right? P: No lah, sometimes I modify if their suggestions are not useful. -Ben’s post-intervention interview, intermediate proficiency boy

The interview excerpts of Luke and Ben revealed that a key strategy that both the boys called on was to ask for feedback from the people around them. They also exhibited how they handled such feedback. Such a finding demonstrates the important role that the socio-cultural dimension of learning (e.g., social mediations) plays in the pupils’ learning environment. The qualitative analyses of the student interviews, some of which have been presented above, show that the experimental pupils, regardless of their proficiency levels, were able to do planning more extensively as compared with their pre-intervention planning. They were also better in orchestrating their strategies. For example, Jia Min not only reported planning the content, e.g., the characters, the events and how the events took place, but also reported assessing the audience in terms of choosing the right words. The qualitative analyses suggested that the participants realized the effectiveness of strategy use and tried using the strategies they were taught during the intervention. Specifically, the findings demonstrated that the SBI was successful in helping the students 1) gain awareness of the effectiveness of their use of existing strategies, 2) expand their repertoire of writing strategies, 3) understand the usefulness of the

Chapter Six

174

writing strategies learned during the SBI, and 4) better orchestrate their writing strategies. Such findings prove that the SBI resulted in the gains of the participants’ strategy use both quantitatively and qualitatively.

6.2.2 Teacher interviews While detailed analyses of teacher interviews go beyond the scope of this study, it is still useful to find out what the experimental teachers thought about the intervention and what changes the SBI brought about in them. By analyzing some teacher interviews, more insights on the effects of the intervention can be gained. In this section, I will focus on the interview with one teacher, who shared with me her convictions on the intervention. Representative of all the six experimental teachers who were involved in the SBI, Madam F was also willing to share her opinions in many occasions. When asked about her observation on her students’ writing after the intervention, she expressed her affirmation of the positive effects of the intervention (see Excerpt 17). Excerpt 17 Yeah, they have a guideline in their head to follow through, you see, so their compo is longer this time, because they’ve more ideas. Then they... they can write better in terms of the adjectives and the choice of words and they can develop the story more lah. So I find that it is good to teach them the strategies. I used to teach them the strategies but it’s ad-hoc. It’s not as structured as your lesson plans. So there’s no linkage in between the strategies that I’ve taught them, so they don’t know how to apply. But this time it’s very structured, so we follow the steps. I have some pupils who really, you know, consciously, really do it and then they find that they benefit. Madam F

She felt her students had a guideline in writing their compositions. They could write longer text as they had more ideas and vocabulary at their disposal. She did teach her students some writing strategies before but in an ad hoc manner. Her students did not know how to apply those strategies to their writing, whereas the SBI lesson plans helped her teach better. She also felt that some of her students applied the writing strategies consciously. She was very happy to be involved in the project during which she could teach her students useful writing strategies (see Excerpt 18). In a way, the

Findings and Discussion of Phase Two

175

project gave her the opportunity to try these strategies systematically. Otherwise, she would not have attempted to do it. Excerpt 18 Yeah, I’m actually quite happy to be involved in this project, because it gives me a chance to uh, really, teach my class useful strategies for them in their writing, because they are actually very weak. So the weekly lesson actually forced me... to really teach them. Otherwise I will be like trying to postpone it until the time I am free... which I will never be free... yeah so in a way it’s good la, in a way. Madam F

At the same time, the SBI was also a learning process for her as she mentioned “…helps me also in my professional development” (see Excerpt 19). As a matter of fact, all the experimental teachers expressed to me their concern over the time that they would have to put in the project at the beginning of the intervention. However, I was very happy to learn towards the end of the project that they all felt that the SBI helped them with their professional development in teaching of writing. Indeed, as was pointed out by O’Malley and Chamot (1990), classroom teachers should be equipped with necessary instructional skills in order for the intervention to be successful. Excerpt 19 I must not think that it is a waste of time, because since I’m involved I may as well do it, that’s my mentality... so I want them to learn... in fact I also learn. Because I find that it helps me also in my professional development, in the sense that I have now this PowerPoint with me. It is not that I do it for you. I have it as my own. Then I can always use it again, you see. Madam F

6.2.3 Conferencing, observations and journal logs The analyses of some of the research measures such as conferencing, researcher observations, and teacher journal logs also indicated that the intervention was conducted successfully. As the researcher, I had weekly conferencing with the class teachers in the experimental group prior to and after each lesson and teacher interviews before and after the intervention. I shared my feedback and suggestions on the teachers’ performance and went through the lesson plan for the following lesson with them in our weekly meetings. The weekly conferencing sessions were very useful and

176

Chapter Six

conducive in that they provided a venue for collective reflection and support. I also went through the teacher journal logs submitted to me after each lesson. Through the teacher journal logs, I understood the teachers better and could communicate them well during the SBI. For example, one teacher wrote in her first journal jog that she found the topic alien and was not confident in teaching the lesson well to the pupils. She felt her pupils may not understand lesson well. Immediately in the following conferencing session, I gave my feedback on her teaching and the teacher gained more confidence gradually. Then, she recorded her positive feelings, “Majority of the pupils enjoyed the lesson and found the strategy useful. A significant percentage of the pupils indicated that they will use the strategy during their writing.” Similar positive notes were also found in her subsequent journal logs. Indeed, the reflection, sharing, monitoring and feedback during the intervention may well help make the SBI successful. My observations also showed that the teachers liked the SBI package as the materials were co-developed by the teachers during the SBI pilot, who understood their colleagues, i.e., teachers from other primary schools. All the materials were made readily available in the SBI package, which included the lesson plans, worksheets, PowerPoint slides and so on. Some of the teachers reflected during our interactions they could better use the time saved from preparing the materials to understand and prepare for their teaching. As commended by one of the teachers, the SBI package was like a “Bible” for the writing teachers.

6.3 Conclusion The strategy-based writing instruction was successfully implemented in a natural teaching environment in Singapore primary schools. There were several kinds of evidence showing that the SBI was effective. First, the SBI adopted a mixed-methods approach to triangulate the data: both quantitative and qualitative data were collected (Dörnyei, 2007; Mackey & Gass, 2005). The quantitative results (see 6.1) suggested that the SBI was successful in improving both the pupils’ writing competence and their strategy use. However, it was made clear in chapter 4 that the quantitative data collected through the questionnaire could only show the frequency of strategy use and could not fully capture the complex interactions between strategy use and the pupils’ English proficiency, e.g., orchestration of

Findings and Discussion of Phase Two

177

strategy use. The experimental impact resulted from the SBI could not be fully understood or appreciated without the qualitative data. Therefore, qualitative data were also collected in order to better understand the effectiveness of the SBI. The qualitative results (see 6.2) also provided robust evidence showing that the SBI was successful. It was found that the pupils became more aware of their strategy use, gained a larger repertoire of LLS, and orchestrated their strategies more effectively. A mixedmethods approach offers the potential for deeper understandings of the SBI. It is also important to highlight that the effects of the SBI, though with some decline, were sustained without subsequent exposure as shown by the delayed test. In other words, another possible confounding factor, increased writing practice, has been ruled out. Overall, both the quantitative and qualitative analyses showed positive impact of the SBI on the experimental group. Furthermore, both the experimental and the control groups received the same amount of writing instruction from their own writing teachers during the intervention period. The experimental teachers used the SBI package whereas the control teachers carried out their teaching of writing as planned by the schools. The only major difference between the two groups lay in the SBI. Therefore, it is highly unlikely that the pupils improved both in their writing competence and use of the target strategies due to factors other than the intervention.

CHAPTER SEVEN CONCLUSIONS AND IMPLICATIONS

The present study examined four research questions concerning the writing strategies used by primary school pupils and the effectiveness of strategy-based writing instruction in Singapore. The research questions were: 1) What writing strategies do upper primary school pupils use? 2) How are the differences in strategy use related to language competence? 3) In what ways does strategy-based writing instruction help pupils improve their writing competence in English? 4) In what ways does such instruction help improve pupils’ use of the target writing strategies? To address these questions, I collected both quantitative and qualitative data by using interviews, a writing strategy questionnaire designed for the present study, students’ end-of-semester English exam results, pre-test, post-test, delayed test, classroom observations, conferencing and teacher journal logs. I presented the research methodology in chapter 4 and reported the results in chapter 5 (Phase 1), and chapter 6 (Phase 2), respectively. In this chapter, I will begin with a summary of the major findings. Then, I will draw implications in light of LLS theories and pedagogies. Finally, I will discuss limitations of the present study and make relevant suggestions for future research

7.1 Summary of major findings The results presented in the preceding chapters have shown robust patterns of writing strategy use and consistent relationships between strategy use and language learning achievements across the two phases of the research project. This section will be highlighting some of the most important findings that I believe are most relevant to Singapore primary school pupils.

Conclusions and Implications

179

7.1.1 Major findings from Phase one The findings of Phase 1 showed a significant relationship between strategy use and English learning achievements on a large scale through a questionnaire survey with more than 3,000 upper primary school pupils. This phase involved five primary schools in Singapore, with each being analyzed individually first and then taken together to inform the design of Phase 2. If similar/consistent patterns between strategy use and language learning outcomes could be established across the five participating schools, these would provide robust evidence of the observed relationships. . In the quantitative analyses, the planning strategies, feedback handling strategies and revising strategies were consistently found to have a significant positive relationship with English language learning outcomes in four of the five schools. In addition, text-generating strategies were also found to predict the English language learning outcomes well in all the five schools. The overall pattern is that the more the pupils reported employing these four groups of writing strategies, the better English language learning outcomes they had. However, no significant differences were found between the high proficiency and the low proficiency pupils in their use of self-initiating strategies, monitoring and evaluating strategies, resourcing strategies and social/affective strategies. The findings lend support to the importance of planning, text-generating and revising strategies in the LLS literature (Chamot & El-Dinary, 1999; De Silva, 2015; Gu et al., 2005; O’Malley et al., 1985a; O’Malley & Chamot, 1990; Vandergrift, 2003a; Victori, 1999). Therefore, these groups of writing strategies should be taught during the strategy-based writing instruction.

7.1.2 Major findings from Phase two Phase 2 was an intervention study, which focused on the development of the participants’ writing competence as well as strategy use, involving 442 pupils, with a relatively large sample size as compared with other similar interventions. The participants were primary 5 pupils from three primary schools in Singapore. In terms of writing competence, the experimental group made significant gains on the post-test, with an effect size (Ș2 = 0.118) close to a large effect (Ș2 = 0.14). The treatment effect was found to have declined on the delayed test to a moderate effect size (Ș2 = 0.06). Nonetheless, the analyses of the delayed test still showed evidence of

180

Chapter Seven

retention of the gains in writing competence with the passage of time and without subsequent writing practice (cf. Macaro, 2009). The quantitative analyses showed that a significant treatment effect was also found for three of the four target groups of strategies taught in the intervention, i.e., text-generating strategies, feedback handling strategies, and revising strategies. However, the effect size fell into the range from .010 to .020, indicating a small effect size. There are at least two reasons why only a small effect size was obtained on the target writing strategies. First, the individual strategies of each group of writing strategies taught in the intervention only represent a subset of the strategies for that group. The pupils may have improved in their use of the taught strategies, but not in those that were not taught. As a result, their improvement was not substantial enough to greatly improve the overall score of the subscale in the questionnaire. Second, the questionnaire only elicited frequency data on strategy use. However, as has been noted previously, effective strategy use does not only entail more frequent use of LLS but also orchestration of strategy use, which the questionnaire could not capture. Taken together, the findings do yield robust evidence suggesting that the SBI was successful in helping improve the participants’ strategy use. As regards planning strategies, the SBI yielded no significant treatment effect. It is contended that the finding is not disappointing for three reasons. First, only half of the 10 strategies that appeared in the questionnaire were actually taught during the intervention. Second, only frequency data was captured by the questionnaire, instead of the quality of strategy use, i.e., orchestration. The qualitative analyses of the student interviews did reveal better use of planning strategies for the participants in the experimental group (see section 6.2 for details). In addition, the qualitative analyses of Phase 2 also showed robust evidence that the intervention was successful in helping the pupils improve their strategy use qualitatively. The pupils became more aware of how to employ their existing writing strategies more effectively, expanded their repertoire of writing strategies, and orchestrated their strategy use qualitatively better after the intervention. On the whole, the SBI has produced evidence for a causal link between strategy use and language learning outcomes. The writing intervention has thus produced solid empirical evidence that strategy-based writing

Conclusions and Implications

181

instruction can foster Singaporean pupils’ writing competence and help improve their use of writing strategies.

7.2 Theoretical implications The study has drawn on several learning theories, i.e., Anderson’s Adaptive Control of Thought (e.g., 1976, 1982, 1983, 1985, 1992, and 1993), Vygotsky’s (1978, 1986) social mediations and zone of proximal development, Feuerstein’s (1990) theory of structural cognitive modifiability and mediated learning experience, as well as Kucer’s (2001, 2005, 2009) multi-dimensional view of literacy, in order to ground the present study on a sound theoretical base. Although the study does not purport to test the soundness of these theories with respect to LLS research directly, they nonetheless offer a useful conceptual framework for the design of similar studies. In particular, the present study represents an effort to integrate different theoretical perspectives, i.e., cognitive and socio-cultural, into one general framework. Traditionally, the majority of LLS studies have relied heavily on cognitive theories, e.g., Anderson’s ACT, in viewing learning strategies. The cognitive theories can offer great potentials in explaining what language learning strategies are, and how they are learned and taught. However, the cognitive strategies are fraught with limitations with regard to LLS research (see section 3.1). Therefore, the present study also drew on some socio-cultural theories to complement the inadequacy that arose from the cognitive theories. From the socio-cultural perspectives, the teacher-student interactions/mediations play a very important role in the development of students’ learning strategies (Coyle, 2007). Young learners spend most of their time in the classroom where teachers mediate their learning. The teacher-student interactions constitute most of students’ learning time in the classroom. There has been a growing trend in LLS research incorporating socio-cultural theories (e.g., Coyle, 2007; Donato & McCormick, 1994; Guerrero & Villamil, 2000; Hu, 2005). Second language researchers have also called for the union of socio-cultural theories and cognitive linguistics (e.g., Lantolf, 2000, 2006; Lantolf & Thorne, 2006). Phase 2 of the study has demonstrated successfully how socio-cultural theories could be implemented in the design of the SBI. For example, mediated learning experience (Feuerstein, 1990) was an important concept that was conveyed to the experimental teachers during

182

Chapter Seven

the workshop and conferencing sessions. They were trained to incorporate these theories into their teaching. Vygotsky’s (1978, 1986) zone of proximal development was embedded in the lesson plans in such a way that at the beginning of each lesson (presentation) the writing teachers assumed most of the responsibility, but eventually the teachers released it and helped the pupils take up more responsibility until they became independent at the end of the lesson (expansion). A unique contribution of the present study is to situate LLS research in the multi-dimensional view. Specifically, the present study posits that literacy learning takes place in multiple dimensions, i.e., cognitive, linguistic, socio-cultural, and developmental dimensions (Kucer, 2001, 2005, 2009), so do LLS. Besides the cognitive and socio-cultural dimensions (e.g., Lantolf, 2000, 2006; Lantolf & Thorne, 2006), the present study has incorporated the linguistic dimension into the model of LLS development. For most of LLS researchers (e.g., Coyle, 2007; Donato & McCormick, 1994; Guerrero & Villamil, 2000; Hu, 2005), language is only an important tool that mediates the development of LLS. However, the present study not only perceives language as an essential tool but also as an important product as a result of LLS development. Given the complexity, the present project was not to collect data to validate such a theory. Rather, the multi-dimensional view was employed to help explain what LLS are and how LLS can be learned and taught. In the cognitive dimension, writers have such writing strategies as planning, monitoring, generating text, and revising at their disposal. These writing strategies are perceived as cognitive processes or mechanisms to help learners write more effectively. Results from the two phases (see chapters 5 & 6) have illustrated clearly that the participants used a wide range of writing strategies. Successful writers and their unsuccessful counterparts employed strategies differently. The results suggested that the pupils reported using their strategies effectively both quantitatively and qualitatively as a result of the SBI. With respect to the linguistic dimension, writers need to possess the necessary linguistic competence or knowledge to use certain strategies. In turn, their improved strategy use also contributes to the development of their linguistic competence. The writers’ linguistic knowledge of LLS may promote their use of LLS. Language is a primary mediation tool, through which the learning of LLS can take place (see section 3.3). Therefore, the

Conclusions and Implications

183

extent to which the writers understand what LLS are, how they are used and why they are used can well demonstrate how the writers may progress through SBI.

Figure 7.1. Dimensions of language learning strategies

The qualitative analyses of student interviews of Phase 2 show that both the high and the low proficiency pupils improved on the linguistic dimension of LLS after the intervention. As is evident in the qualitative analyses of Phase 2, the pupils made considerable progress in their linguistic repertoire of LLS; they were able to use the strategy vocabulary learned during the SBI to discuss their strategy use. For example, Angelina talked about overall goals, specific goals, organization, the audience, settings, which she did not know at all before the SBI. The other pupils also reported using planning and what is involved in planning to a certain degree in addition to some other writing strategies taught during the intervention. Of course, such analysis may not show conclusive evidence about a significant difference in the participants’ overall strategy use before and after the SBI. However, it does suggest that the intervention brought about positive change in their linguistic knowledge about writing strategies. “Language is the most powerful tool which mediates human development, and it is fundamental to the effectiveness of learning.”

184

Chapter Seven

(Coyle, 2007, p. 66) As has been discussed previously, writers’ linguistic knowledge of LLS may promote their strategy use when language is perceived as a mediation tool by which the learning of LLS can take place. Therefore, the findings have demonstrated from the linguistic dimension that the SBI was successful in improving the participants’ knowledge of writing strategies. The present study also draws attention to the importance of the sociocultural dimension in LLS research. As was theorized in sections 3.2 and 3.3, learners acquire their LLS through the social interactions/mediations that they are engaged in with their teachers or other learners. In other words, the socio-cultural dimension enables us to understand the learning of LLS in the actual context, i.e., teacher-learner interactions in the classroom (Walters, 2007). Specifically, the present study adopted Feuerstein’s (1990) mediated learning experience (MLE) in guiding the design of Phase 2, which sets principles on how the teacher-learner interactions should be actualized. The three principles of MLE, i.e., intentionality, mediation of meaning and transcendence were weaved into the lesson plans for the teachers in the experimental group. Training sessions were provided for them to understand and learn what MLE entailed. I, as both the trainer and the observer, provided a conferencing session before every lesson for the experimental teachers to grasp the ideas in the lesson plans followed by my observation of their teaching and a feedback session afterwards to ensure that the MLE principles were actualized. To achieve intentionality, the teachers were instructed to explain what feedback handling was about, what was involved, and how feedback handling could be carried out in the classroom. The teachers were also asked to choose pupils to act out how an appropriate feedback handling session should work out for the class. For example, as was required in Lesson 9, the teacher “should demonstrate how to provide polite, constructive and helpful suggestions and comments”. The teachers would perform explicit thinking aloud for the class to understand how this process could be done. In terms of mediation of meaning, the class was given a task to apply what they had just learned. They would then reflect on whether they employed the target strategies, i.e., feedback handling strategies, how they were used, and whether the feedback handling strategies were useful to help them write more effectively. Transcendence was achieved by giving the pupils another writing task where feedback

Conclusions and Implications

185

handling was involved so that they saw the importance of such strategies in improving their writing competence. Besides the social mediations between the teacher and the pupils, the pupils’ peers were also found to mediate on their writing processes. The research findings from the qualitative analyses of Phase 2 suggested that the writing environment (e.g., classroom and home), the social interactions between learners, between learners and their teachers, and between learners and their family members were also key in helping them improve their writing. Such findings support the socio-cultural dimension view on learning and LLS in that the two are essentially social and take place in social mediations/interactions (Coyle, 2007; Donato & McCormick, 2004; Walters, 2007). As regards the developmental dimension, the current study has successfully demonstrated the growth of the other three dimensions (linguistic, cognitive, and socio-cultural) through the writers’ use of LLS. The three dimensions transacted with each other in a complex way. For example, it was found in the two phases that the high proficiency pupils showed better development in the cognitive dimension as they reported using more strategies than their low proficiency counterparts. At the same time, they possessed a better linguistic repertoire than those low proficiency pupils. Likewise, Phase 2 found that after the SBI, the participants in the experimental group were better able to report their strategy use and also employed more writing strategies. Specifically, the present study showed how the experimental teachers demonstrated to the pupils how the learning of LLS occurred in the three dimensions. For example, they thought aloud for the pupils as to what LLS were and how they worked. The teachers also promoted the use of many strategy words with the pupils, e.g., writing strategies, planning, setting goals, and handling feedback. Equipped with these, the pupils could better employ their writing strategies. In the socio-cultural dimension, the teachers mediated on the pupils’ learning of the writing strategies through scaffolding in each lesson. Vygotsky’s (1981) ZPD and Feuerstein’s SCM were applied to the design and implementation of the intervention. Taken together, the present study represents a good effort in incorporating the multiple dimensions which contribute to the pupils’ learning, development and employment of LLS.

186

Chapter Seven

7.3 Pedagogical implications This study is one of the very few two-phase studies that have researched the patterns of strategy use and effectiveness of strategy-based writing instruction with a mixed-methods approach and a large sample size. Given the robust research evidence on the relationships between strategy use and writing competence as well as the effectiveness of the SBI, several pedagogical implications can be drawn. It is high time that strategy-based writing instruction was integrated into regular writing instruction. I will discuss the pedagogical implications in three respects. The first is how teachers can assess their students’ strategy use. The second implication has to do with the kind of SBI that teachers can conduct for their students. Third, I will discuss how teacher training can impact on or mediate the success of interventions on students’ strategy use. First, as the findings suggest, there are many significant differences between the high and the low proficiency pupils in terms of the frequency and orchestration of their strategy use, writing teachers can first find out what the differences are among the pupils they are teaching. Then, they can train a few proficient writers to think aloud before the rest of the class to demonstrate how they use their writing strategies. Young learners are often interested to try using the strategies that their peers suggest as they understand how their peers become good writers through the same activities (Anderson, 2008). Alternatively, writing teachers can pair up proficient writers and their less proficient peers to carry out a writing task together in which the proficient writers should also think aloud so that the less proficient ones can benefit. Teachers can also ask some proficient writers to complete a writing task, and then have a discussion with the class on how writing strategies have been used by the pupils. In any case, proficient writers should be trained on how to do think-alouds appropriately. While proficient writers differ from weak writers in their strategy use, there is also individual variation in strategy use among proficient writers. The same can be true of weak writers. Such inferences serve as reminders to writing teachers that detailed analyses of weak writers’ writing problems are also highly necessary before teachers decide what strategies from those proficient writers will be put on their teaching schedule to teach weak writers. In other words, teachers should also work to increase their awareness of students’ strategy use and needs (Griffiths & Parr, 2001).

Conclusions and Implications

187

Second, the methods that can be used to find out the writing problems or strategies that students may need to learn include, but are not limited to, surveys, interviews, and classroom observations (see Chamot, 2004, 2008; Cohen, 1998; O’Malley & Chamot, 1990; Oxford, 1990). Teachers can develop a questionnaire suitable for their students’ proficiency level to find out the general picture of students’ use of writing strategies. Questionnaires allow teachers to collect data from large numbers of students, which helps save a lot of time. The data collected through the questionnaire can be correlated with students’ age, gender and proficiency level and so on for teachers to analyze how students use their strategies and what their problems are. Moreover, questionnaires followed by classroom discussions on strategy use may help students orchestrate their strategies more effectively because such discussions help raise students’ awareness on how their strategies have been used (Anderson, 2008). Available questionnaires include the writing strategy questionnaire developed in Phase 1 of this study and others that can be found in the literature (e.g., Bedell & Oxford, 1996; Cohen & Weaver, 1998; Cohen et al., 1998; Griffiths, 2007; Oxford, 1990; Petriü & Czárl, 2003; Wharton, 2000). One caution with use of the questionnaires developed in other cultures is that those questionnaires may not inform teachers or researchers of what learners think in their local culture (Bremner, 1998; Petriü & Czárl, 2003; Zhang, 2003). The writing strategy questionnaire developed in the present study may be a more suitable instrument for primary school pupils, teachers and researchers in Singapore given the grounded approach involved in its development (Macaro & Erler, 2008). One common issue with all the above-mentioned data collection methods is that most of them are exclusive; each method tends to focus only on certain aspects of LLS, which offers only partial information on learners’ strategy use (Rao, 2002). Therefore, I recommend triangulation of the data collection methods with the same sample of students. This can enhance the validity of the research methods. For example, it would be useful to measure the correlations of the results collected from a questionnaire survey, think-aloud protocols, interviews, and classroom observations. A high correlation between these methods would suggest a higher accuracy of the data collected using these methods. Of course, to realize triangulation of data collection methods, writing teachers may need proper training or have to collaborate with researchers. Third, writing teachers should help students expand their strategy repertoires and encourage them to employ strategies more often. The

188

Chapter Seven

findings of Phase 1 suggest that the high proficiency pupils have a larger repertoire of writing strategies at their disposal and use them more often than their low proficiency peers. Phase 2 established a causal relationship between strategy use and writing competence. Overall, the more strategies students use and the more often students use them, the better writing competence they can achieve. With such robust research evidence, writing teachers should have the confidence to help expand students’ repertoires of writing strategies and encourage them to use their strategies more often, instead of always spending a substantial period of time in every writing lesson explaining the usage of useful words and phrases that students should put in their compositions. Fourth, the success of Phase 2 suggests that strategy-based writing instruction should be conducted explicitly and integrated into the writing class. The explicit writing strategy instruction essentially involved five recognizable procedures: 1) presentation, i.e., raising students’ awareness of their strategy use; 2) presentation, i.e., teacher modeling of strategy use; 3) practice, i.e., student practice with new strategies; 4) evaluation, i.e., evaluation of strategies used; and 5) expansion, i.e., transferring strategies to new tasks. Most researchers have appeared to agree to the five-procedure model for strategy instruction (Chamot, 2008; Chamot et al., 1999; Macaro, 2008). Writing teacher’s responsibility is reduced along the process where students take more responsibility. The goal of the fiveprocedure model is to help students to use the writing strategies taught independently as well as in new tasks. The carefully designed fiveprocedure model with teachers’ input serves as an important instructional tool for strategy-based writing instruction. The intervention is designed and conducted from the perspective of the participants and the teachers, rather than only from that of the researchers (Chamot, 2005). While some researchers may still have concerns over the issue of integration of strategy instruction into language curriculum (see Chamot, 2008; Gu, 1996), I would argue that explicit strategy instruction should be embedded into the writing class, given the robust research evidence produced in Phase 2. The reasons are obvious. First, students will have opportunities to use writing strategies with authentic writing tasks. Second, such a writing class also provides the writing teacher with the opportunities to see how his/her students employ the strategies and give his/her feedback on the students’ strategy use.

Conclusions and Implications

189

When introducing strategies, teachers should provide ample examples on the conditions in which a strategy or more than one strategy (orchestration of strategy use) can be used effectively. Young students should be given opportunities to see and experience how, when and why these strategies are employed in class (Macaro, 2007). For example, a recurring focal point of the lesson plans was to make the pupils understand what the target strategies were, when they were normally used, and why they were used. These are important points to include in strategy instruction as only concrete examples, rather than abstract ones, can help students learn effectively. Fifth, in tandem with the structure of the SBI lesson plans, writing teachers should design their intervention based on a plan-draft-revise model. As has been reviewed previously, planning, writing/translating (text-generating) and revising are the three most important processes in writing instruction (Flower & Hayes, 1981; Graham, 2006a). Effective writing instruction should always involve the teaching of the three important writing processes (see Harris et al., 2010; Pressley et al., 2007). Sixth, another implication from the findings of Phase 2 concerns the professional development of teachers, who will eventually become experts in teaching writing strategies through the SBI. To my best knowledge, very few LLS studies (e.g., Chamot et al., 1999) have paid attention to the development of strategic teachers. In line with what Chamot (2005) rightfully points out, “expertise in teaching language learning strategies must be integrated into pre- and in-service preparation if teachers are to provide it to their L2 students” (p. 123), I would argue that strategy instruction should not only be looking to help students become strategic learners but also help teachers become expert in strategy instruction. This is also to respond to some LLS researchers’ call for more professional development in strategy instruction for regular classroom teachers (e.g., Shapira & Hertz-Lazarowitz, 2005). Likewise, Graham and Perin (2007) note that there was a better effect on students’ writing when teachers were trained in strategy instruction. Seventh, the present study provides important insights on how teacher training should be conducted to help teachers develop a solid understanding of the nature of LLS and SBI. In this study, a 3-hour SBI workshop was conducted to train the teachers before the intervention. During the intervention, I had weekly conferencing sessions with all the experimental teachers, in which feedback and suggestions were provided.

190

Chapter Seven

The teachers and I would go through the lesson plan for the following week together. If the teachers had any questions, I would also help answer them either during a conferencing session or via email. In addition, some journal articles on SBI research were also provided to the teachers for them to understand more about LLS in general and SBI in particular. Taken together, all these have contributed to the success of the intervention. Therefore, I hold the view that strategic teachers are an indispensable part of successful strategy intervention and call for more research involving teachers as an important variable.

7.4 Limitations of the study There may be limitations with regard to the writing strategy questionnaire developed in Phase 1. As has been noted in the literature (see Chamot, 2004; 2008), when responding to the questionnaire, the participants may have forgotten the strategies they employed in the past; they may claim to use strategies they have never used before; or they may misunderstand the questions on the questionnaire. For these reasons, Phase 2, which involved the same writing strategy questionnaire, also adopted other data collection methods, such as student interviews, teacher interviews, and classroom observations. It was already made clear previously, a combination of these research methods would help enhance the quality of data. Another limitation of the questionnaire is the relatively low alphas for three of the subscales. Feedback handling, resourcing, and social/affective strategies all had an alpha value below .60. Although such low alphas were still within the range of acceptability (DeVellis 1991; Dörnyei, 2007; Petriü & Czárl, 2003), they may pose a threat to the internal consistency of the questionnaire to a certain extent. A possible reason may lie in the small number of items that these subscales included. This was particularly evident in the case of resourcing strategies, which only included three items. In order to minimize this problem, the study went through five cycles, developing the questionnaire. Such a measure was to follow Petriü and Czárl’s (2003) suggestion on how to enhance the overall validity of the questionnaire. The questionnaire was first designed by a group of experienced LLS researchers in light of the available LLS literature. Then, amendments were made based on the feedback and suggestions of some local primary school teachers. Primary school pupils’ feedback was also sought to further validate the questionnaire. Finally, the questionnaire was piloted with 221 pupils in a local school. All these measures helped

Conclusions and Implications

191

improve the content validity of the questionnaire in particular and the overall validity in general. The sample may lack representativeness to the whole population of primary school pupils in Singapore in that the participants of Phase 1 were upper primary school pupils, i.e., primary 4-6. Phase 1 involved a relatively large sample of more than 3,000 pupils. However, no lower primary pupils were involved. This is a limitation that prevents generalization of the findings to the population of lower primary pupils in Singapore. Likewise, lack of randomization of the participants in Phase 2 can be seen as another limitation. All the participants from the six experimental and six control classes stayed in their own intact classes throughout the entire intervention period. One could argue that random assignment of participants is a necessary requirement of a rigorous experimental design. However, random assignment of participants for the present study was not possible and the method for assigning classes to the experimental or control groups was the best that could be done given the situation in the schools. The SBI lasted 11 weeks (80% of the time of a semester) and involved four classes in each of the three schools. As all the classes in each school had different timetables, random assignment of participants would necessitate changes in their timetables, which, as a result, would cause many logistical issues to the schools. It may be that a Hawthorne effect can account for part of the gains on the participants’ writing competence and strategies in Phase 2. That is, simply being part of the intervention and participating in the SBI prompted the participants to study in more motivated ways, thus improving their writing competence and strategy use. The present study addressed this issue by assessing the non-target strategies on which the SBI should have yielded less or no effect. The rationale for inclusion of these non-target strategies was that if no significant gains on these strategies were obtained, this would rule out the possibility that mere participation rather than the intervention per se had yielded effects. It was found that the intervention yielded no effect on the four non-target groups of writing strategies. Besides, although measures were taken to ensure that the experimental and the control groups were as comparable as possible, better classes served as the control group in all the three participating schools, which may disadvantage the experimental group with regard to analyses of the effects of the intervention. In the Singapore context, teachers are very competitive.

192

Chapter Seven

The control teachers, who were well aware of the SBI and may be worried about the possibility that their classes might fall behind the experimental classes due to the SBI, tried very hard to beat the experimental classes. As my observations in the control classes also revealed, the pupils in these classes knew that they were part of the research project as well. Taken together, I would argue that the Hawthorn effect has been less of a concern for the above-mentioned reasons. Diffusion effects (Craven, Marsh, Debus, & Jayasinghe, 2001) may constitute another limitation of the study. A diffusion effect may take place when there are direct or indirect interactions between the experimental and the control groups during the intervention. In such interactions, the control group may learn about the information intended for the experimental group. As a result, they would try their best to compete with the experimental group. Another diffusion effect happens when the control group is aware that the experimental group is receiving an intervention or “better treatment”, which may result in resentful demoralization, in which the control group may give up or stop trying. It was not possible to prevent these diffusion effects from taking place during the intervention as the pupils from both the experimental and the control groups interacted freely during the period. However, these effects were less of a concern because there were minimal gains on the writing competence of the control group at both the post- and delayed tests. In addition, the control group stayed predominantly unchanged on the target strategies on the post-intervention questionnaire.

7.5 Recommendations for future research This study aimed at identifying patterns of strategy use as well finding the effectiveness of strategy-based writing instruction in the Singapore context. While replication studies are needed, future intervention research should also explore systematically the relationship between effective strategy instruction and teacher characteristics, which may include, but are not limited to, teacher attitude, amount of LLS intervention training, teacher involvement, and years of teaching experience. As has been pointed out previously, teachers are a very important variable in the classroom where learning takes place. However, very little attention has been paid to this important variable of LLS research so far. More intervention studies should be undertaken to evaluate which teacher characteristics may mediate the effectiveness of strategy instruction and to what extent.

Conclusions and Implications

193

In terms of research design, the LLS research field would benefit from more longitudinal studies. The effect size of the 11-week intervention declined at the delayed test, which was administered in the 15th week. To my knowledge, very few studies have adopted a longitudinal design (e.g., Fidalgo et al., 2008; Mizumoto & Takeuchi, 2009). More longitudinal studies would help ascertain the long-term effect of strategy-based writing instruction. With respect to the participants of Phase 2, this study only involved Singapore primary 5 pupils. While the positive results of the intervention can help inform primary school teachers in their teaching of writing, it is not clear whether the same intervention on the other upper primary school grade levels, i.e., primary 4 and primary 6, will achieve the same results. Cross-sectional studies involving more upper primary grade levels may offer more insights and help draw more definite conclusions on the effect of strategy-based instruction. After sufficient knowledge on writing strategies is obtained from the present study as well as those aforementioned studies, I would like to see more of both descriptive and intervention studies with regard to writing strategies. In particular, we are now better equipped with theoretical and empirical findings from this study and others in the LLS literature. LLS researchers are now in a better position to formulate better instructional models of writing strategies for the purpose of developing strategic learners as well as teachers. With the proposed theoretical framework and research methodologies of the present study, more research studies should be undertaken to gather empirical data to test the applicability of the framework. This study serves as a starting point towards developing standardized ways of investigating the relationship between strategy use and language learning outcomes as well as the effectiveness of strategy-based instruction.

7.6 Conclusion To conclude, this study has investigated the relationship between language learning strategies and language learning outcomes. In addition, it has also examined the effectiveness of strategy-based writing instruction in Singapore primary schools. The results are very encouraging as the findings have shown that there is a strong relationship between the use of LLS and language learning outcomes. Effective writers used more writing

194

Chapter Seven

strategies and effectively orchestrated their strategy use as compared with their ineffective counterparts. Therefore, the causality between strategy use and language learning outcomes has been established. The success of the intervention has indicated that strategy-based writing instruction has produced definite positive results in Singapore primary schools. Therefore, strategy-based writing instruction is a viable approach to teaching writing in Singapore, and possibly other countries, with many more benefits than the traditional approach to teaching writing. The longitudinal design is also one of a kind in the LLS research field that has involved two main phases, with the first phase having different yet necessary objectives that worked as an indispensable precursor to the second phase. The present study has also shown that the LLS research can be undertaken following multi-dimensional views theoretically, thus challenging research that has focused only on a unitary dimension such as the cognitive or socio-cultural dimension.

REFERENCES

Afflerbach, P., Pearson, P. D., & Paris, S. G (2008). Clarifying differences between reading strategies and reading skills. The Reading Teacher, 61(1), 364-373. American Educational Research Association. (2011). Code of Ethics. Retrieved from http://www.aera.net/AboutAERA/AERARulesPolicies/ProfessionalEth ics/tabid/10200/Default.aspx Anderson, J. R. (1976). Language, memory, and thought. Hillsdale, NJ: Erlbaum. Anderson, J. R. (1980). Cognitive psychology and its implications. San Francisco: Freeman. Anderson, J. R. (1982). Acquisition of cognitive skills. Psychological Review, 89, 369-406. Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard University Press. Anderson, J. R. (1985). Cognitive psychology and its implications (2nd ed.). New York: Freeman. Anderson, J. R. (1987). Skill acquisition: Compilation of weak-method problem solution. Psychological Review, 94, 192-210. Anderson, J. R. (1990). Cognitive psychology and its implications (3rd ed.). New York: Freeman. Anderson, J. R. (1992). Automaticity and the ACT* theory. American Journal of Psychology, 105, 165-180. Anderson, J. R. (1993). Rules of the mind. Hillsdale, NJ: Erlbaum. Anderson, J. R. (1995). ACT: A simple theory of complex of cognition. American Psychologist, 51, 355-365. Anderson, J. R. (2000). Cognitive psychology and its implications (5th ed.). New York: Freeman. Anderson, J. R. (2005). Cognitive psychology and its implications (6th ed.). New York: Freeman. Anderson, J. R., & Lebiere, C. (1998). The atomic components of thought. Mahwah, NJ: Erlbaum. Anderson, J. R., & Schunn, C. D. (2000). Implications of the ACT-R learning theory: No magic bullets. In R. Glaser (Ed.), Advances in

196

References

instructional psychology: Educational design and cognitive science (Vol. 5, pp. 1-34). Mahwah, NJ: Lawrence Erlbaum Associates. Anderson, N. J. (1991). Individual differences in strategy use in second language reading and testing. The Modern Language Journal, 75, 460472. Anderson, N. J. (2005). L2 learning strategies. In E. Hinkel (Ed.), Handbook of research in second language teaching and learning (pp. 597-613). Mahwah, NJ: Lawrence Erlbaum Associates. Anderson, N. J. (2008). Metacognition and good language learners. In C. Griffiths (Ed.), Lessons from good language learners (pp. 99-109). Cambridge: Cambridge University Press. Anstey, M., & Bull, G. (2006). Teaching and learning multiliteracies: Changing times, changing literacies. Newark, DE: International Reading Association. Ary, D., Jacobs, L. C., Razavieh, A., & Sorensen, C. (2006). Introduction to research in education (7th ed.). Belmont, CA: Thomson Wadsworth. Atkinson, D. (2003a). L2 writing in the post-process era: Introduction. Journal of Second Language Writing, 12(1), 3-15. Atkinson, D. (2003b). Writing and culture in the post-process era. Journal of Second Language Writing, 12(1), 49-63. Bai, B. (2015). An intervention study of writing strategies in Singapore primary schools. System, 53, 96-106. Bai, R. (2005, May). An analysis of writing strategies in ESL composing: A case study of two successful and two unsuccessful writers. Paper presented at the international conference of the Centre for Research in Pedagogy and Practice, National Institute of Education, Nanyang Technological University, Singapore. Bai, R. (2007). Doing research on language learning strategies in Singapore primary schools. SAAL Quarterly, 77, 11-13. Bai, R., & Hu, G. W. (2008, December). A case study of writing strategies used by Singapore pupils. Paper presented at the 13th International Conference on English in Southeast Asia: Englishes and Literatures-in English in a Globalized World, National Institute of Education, Nanyang Technological University, Singapore. Bai, R., & Hu, G.W. (2009, March). A protocol-based case study of good and weak young writers. Paper presented at the TESOL 2009 International Convention, Denver, CO, USA. Bai, R., Hu, G. W., & Gu, Y. (2014). The relationship between writing strategies and English proficiency in Singapore primary schools. The Asia-Pacific Education Researcher, 23, 355-365.

References

197

Baker, W., & Boonkit, K. (2004). Learning strategies in reading and writing: EAP contexts. RELC, 35, 299-328. Bedell, D. A., & Oxford, R. L. (1996). Cross-cultural comparisons of language learning strategies in the People's Republic of China and other countries. In R. L. Oxford (Ed.), Language learning strategies around the world: Cross-cultural perspectives (pp. 47-60). Honolulu, HI: University of Hawaii Press, Second Language Teaching & Curriculum Center. Bereiter, C., & Scardamalia, M. (1987). The psychology of written composition. Hillsdale, NJ: Lawrence Erlbaum Associates. Bitchener, J., Young, S., & Cameron, D. (2005). The effects of different types of corrective feedback on ESL student writing. Journal of Second Language Writing, 14, 191-205. Boscolo, P. (2008). Writing in primary school. In C. Bazerman (Ed.), Handbook of research on writing: History, society, school, individual, text (pp. 293-310). New York: Lawrence Erlbaum Associates. Bremner, S. (1998). Language learning strategies and language proficiency: Investigating the relationship in Hong Kong. Asia Pacific Journal of Language in Education, 1(2), 490-514. Brown, H. D. (2002). Strategies for success: A practical guide to learning English. New York: Addison Wesley Longman. Brown, A. L., Bransford, J. D., Ferrara, R. A., & Campione, J. C. (1983). Learning, remembering and understanding. In J. H. Flavell & M. Markman (Eds.), Carmichael’s manual of child psychology (Vol. 3, pp. 77-166). New York: Wiley. Carr, W., & Kemmis, S. (1986). Becoming critical: Education, knowledge and action research. London: Routledge. Carrell, P. L., Pharis, B. G., & Liberto, J. C. (1989). Metacognitive strategy training for ESL reading. TESOL Quarterly, 23, 647-678. Chamot, A. U. (1987). The learning strategies of ESL students. In A. Wenden & J. Rubin (Eds.), Learner strategies in language learning (pp. 71-83). Englewood Cliffs, NJ: Prentice Hall. Chamot, A. U. (2004). Issues in language learning strategy research and teaching. Electronic Journal of Foreign Language Teaching, 1, 14-26. Chamot, A. U. (2005). Language earning strategies instruction: Current issues and research. Annual Review of Applied Linguistics, 25, 112-130. Chamot, A. U. (2008). Strategy instruction and good language learners. In C. Griffiths (Ed.), Lessons from good language learners (pp. 266-281). Cambridge, UK: Cambridge University Press.

198

References

Chamot, A. U., Barnhardt, S., El-Dinary, P. B., & Robins, J. (1999). The learning strategies handbook. White Plains, NY: Addison Wesley Longman. Chamot, A. U., & El-Dinary, P. B. (1999). Children's learning strategies in language immersion classrooms. The Modern Language Journal, 83, 319-338. Chamot, A. U., & Küpper, L. (1989). Learning strategies in a foreign language institution. Foreign Language Annals, 22, 13-24. Chamot, A. U., Küpper, L., & Impink-Hernandez, M. (1988). A study of learning strategies in foreign language instruction: Findings of the longitudinal study. McLean, VA: Interstate Research Associates. Chen, Y. C. (2007). Learning to learn: The impact of strategy training. ELT Journal, 61(1), 20-29. Chien, S.-C. (2012). Students’ use of writing strategies and their English writing achievements in Taiwan. Asia Pacific Journal of Education, 32(1), 93-112. Cohen, A. D. (1998). Strategies in learning and using a second language. London, UK: Longman. Cohen, A. D., & Macaro, E. (Eds.). (2007). Language learner strategies: 30 years of research and practice. Oxford, UK: Oxford University Press. Cohen, A. D., & Weaver, S. (1998). Strategies-based instructions for second language learners. In W. A. Renandya & G. M. Jacobs (eds.), Learners and language learning (pp. 1-25): Anthology Series 39. Singapore: SEAMEO Regional Language Centre. Cohen, A. D., Weaver, S., & Li, T-Y. (1998). The impact of strategiesbased instruction on speaking a foreign language. In A. D. Cohen (Ed.), Strategies in learning and using a second language (pp. 107-156). London, UK: Longman. Coker, D. (2007). Writing instruction for young children: Methods targeting the multiple demands that writers face. In S. Graham, C. A. MacArthur, & J. Fitzgerald (Eds.), Best practices in writing instruction (pp. 101-118). New York: The Guilford Press. Comber, B. (2002). Critical literacy: Maximising children’s investments in school learning. Draft discussion paper presented at the Resource Teachers: Literacy Training Program, Christchurch, NZ. Coyle, D. (2007). Strategic classrooms: Learning communities which nurture the development of learner strategies. Language Learning Journal, 35, 65-79. Craven, R. G., Marsh, H. W., Debus, R. L., & Jayasinghe, U. (2001). Diffusion effects: control group contamination threats to validity of

References

199

teacher-administered interventions. Journal of Educational Psychology. 93, 639-645. De Silva, R. (2015). Writing strategy instruction: Its impact on writing in a second language for academic purposes. Language Teaching Research. 19, 301-323. DeVellis, R. F. (1991). Scale development: Theory and applications. Cambridge, UK: Cambridge University Press. Donato, R., & McCormick, D. (1994). A sociocultural perspective on language learning strategies: The role of mediation. The Modern Language Journal, 78, 453-464. Dörnyei, Z. (2005). The psychology of the language learner. Mahwah, NJ: Lawrence Erlbaum Associates. Dörnyei, Z. (2007). Research methods in applied linguistics: quantitative, qualitative, and mixed methodologies. Oxford, UK: Oxford University Press. Dörnyei, Z., & Skehan, P. (2003). Individual differences in second language learning. In C. J. Doughty & M. H. Long (Eds.), The handbook of second language acquisition (pp. 589-630). Malden: Blackwell Publishing Ltd. Duffy, G. G. (2002). The case for direct explanation of strategies. In C.C. Block & M. Pressley (Eds.), Comprehension instruction: Researchbased best practices (pp. 28-41). New York: Guildford. Ehrman, M. E., Leaver, B. L., & Oxford, R. L. (2003). A brief review of individual differences in second language learning. System, 31, 313330. Ellis, R. (2008). The study of second language acquisition (2nd ed.). Oxford, UK: Oxford University Press. Falik, L. H. (2000). Mediated learning experience and the counseling process. In A. Kozulin & Y. Rand (Eds.), Experience of mediated learning: An impact of Feuerstein’s theory in education and psychology (pp. 309-23). Oxford, UK: Pergamon. Ferris, D., & Roberts, B. (2001). Error feedback in L2 writing classes: How explicit does it need to be?. Journal of Second Language Writing, 10(3), 161-184. Ferris, D., & Hedgcock, J. S. (2005). Teaching ESL composition (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Feuerstein, R. (1990). The theory of structural cognitive modifiability. In B. Presseisen (Ed.), Learning and thinking styles: Classroom interaction (pp. 68-134). Washington, DC: National Education Association.

200

References

Feuerstein, R., Klein, S., & Tannenbaum, A. (Eds.). (1999), Mediated learning experience (MLE): Theoretical, psychological and learning implications. England, UK: Freund Publishing House LTD. Feuerstein, R., Klein, S., & Tannenbaum, A. (2003). Feuerstein’s theory and applied systems: A reader. Jerusalem: ICELP Press. Feuerstein, R., & Rand, Y. (1997). Don’t accept me as I am: Helping retarded performers to excel. Illinois: Skylight Training and Publishing Inc. Feuerstein, R., & Feuerstein, S. (1999). Mediated learning experience: A theoretical review. In R. Feuerstein, S. Klein, & A. Tannenbaum (Eds.), Mediated learning experience (MLE): Theoretical, psychological and learning implications (pp. 3-52). England, UK: Freund Publishing House LTD. Fidalgo, R., Torrance, M., & García, J-N. (2008). The long-term effects of strategy-focussed writing instruction for grade six students. Contemporary Educational Psychology, 33, 672-693. Field, A. P. (2005). Discovering statistics using SPSS (2nd ed.). London, UK: Sage. Fitzgerald, J. (1987). Research on revision in writing. Review of Educational Research, 57(4), 481-506. Flower, L., & Hayes, J. R. (1980). The cognition of discovery: Defining a rhetorical problem. College composition and communication, 31, 2132. Flower, L., & Hayes, J. R. (1981). A cognitive process theory of writing. College composition and communication, 32, 365-87. Gao, X. S. (2003). Changes in Chinese students' learner strategy: Use after arrival in the UK: A qualitative inquiry. In D. Palfreyman & C. R. Smith (Eds.), Learner autonomy across cultures: Language education perspectives (pp.41-57). New York: Palgrave. Gao, X. S. (2007). Has learning strategy research come to an end? A response to Tseng et al. (2006). Applied Linguistics, 28, 615-620. Gagné, R. M. (1985). The conditions of learning and theory of instruction (4th ed.). New York: Holt, Rinehart and Winston. Gee, J. (1990). Social linguistics and literacies: Ideology in discourses. New York: Falmer. Gee, J. (1996). Social linguistics and literacies: Ideology in discourses (2nd ed.). New York: Falmer. Glaser, C., & Brunstein, J. (2007). Improving fourth-grade pupils’ composition skills: Effects of strategy instruction and self-regulation procedures. Journal of Education Psychology, 99(2), 297-310.

References

201

Goh, C. (2005). Second language listening expertise. In K. Johnson (Ed.), Expertise in second language learning and teaching (pp. 64-84). Basingstoke: Palgrave Macmillan. Goh, C. C. M., Zhang, L. J., Ng, C. H., & Koh, G. H. (2005). Knowledge, beliefs and syllabus implementation: A study of English Language teachers in Singapore. Singapore: Nanyang Technological University. Gordon, L. (2008). Writing and good language learners. In C. Griffiths (Ed.), Lessons from good language learners (pp. 244-254). Cambridge, UK: Cambridge University Press. Grabe, W., & Kaplan, R. (1996). Theory and practice of writing. London, UK: Longman. Graham, S. (2006a). Writing. In P. A. Alexander & P. Winne (Eds.), Handbook of educational psychology (2nd ed., pp. 457-478). New York: Macmillan. Graham, S. (2006b). Strategy instruction and the teaching of writing: A meta-analysis. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (pp. 187-207). New York: The Guilford Press. Graham, S., & Harris, K. R. (2007). Best practices in teaching planning. In S. Graham, C. A. MacArthur, & J. Fitzgerald (Eds.), Best practices in writing instruction (pp. 119-140). New York: The Guilford Press. Graham, S., Harris, K. R., & Mason, L. (2005). Improving the writing performance, knowledge, and self-efficacy of struggling young writers: The effects of self-regulated strategy development. Contemporary Educational Psychology, 30, 207-241. Graham, S., MacArthur, C. A., & Fitzgerald, J. (2007). Introduction: Best practices in writing instruction now. In S. Graham, C. A. MacArthur, & J. Fitzgerald (Eds.), Best practices in writing instruction (pp. 1-9). New York: The Guilford Press. Graham, S., & Perin, D. (2007). A meta-analysis of writing instruction for adolescent students. Journal of Educational Psychology, 77, 445-476. Grenfell, M. (2007). Language learner strategy research and modern foreign language learning and teaching. Language Learning Journal, 35(1), 9-22. Grenfell, M., & Harris, V. (1999). Modern languages and learning strategies: In theory and practice. London, UK: Routledge. Griffiths, C. (2003). Patterns of language learning strategies use. System, 31, 367-383. Griffiths, C. (2007). Language learning strategies: Students’ and teachers’ perceptions. ELT Journal, 61(2), 91-99.

202

References

Griffiths, C. (2008). Strategies and good language learners. In C. Griffiths (Ed.), Lessons from good language learners (pp. 83-98). Cambridge, UK: Cambridge University Press. Griffiths, C. (2015). What have we learned from ‘good language learners’? ELT Journal, 69(4), 425-433. Griffiths, C., & Parr, J. M. (2001). Language learning strategies: theory and perception. ELT Journal, 55(3), 247-254. Gu, Y. (1996). Robin Hood in SLA: What has the learning strategy researcher taught us?. Asian Journal of English Language Teaching, 6, 1-29. Gu, Y. (2002). Gender, Academic major, and vocabulary learning strategies of Chinese EFL learners. RELC Journal, 33, 35-54. Gu, Y. (2003). Fine brush and freehand: The vocabulary learning art of two successful Chinese EFL learners. TESOL Quarterly, 37(1), 73-104. Gu, Y. (2005). Vocabulary learning strategies in the Chinese EFL context. Portland, OR: Marshall Cavendish Academic. Gu, Y. (2007). Foreword. In A. Andrew & E. Macaro (Eds.), Language learner strategies: 30 years of research and practice. Oxford, UK: Oxford University Press. Gu, Y. (2012). Learning strategies: Prototypical core and dimensions of variation. Studies in Self-Access Learning Journal, 3(4), 330-356. Gu, Y., Hu, G. W., & Zhang, L. J. (2004). Eliciting learning strategies from lower primary school students in Singapore. In R. Dawson (Ed.), Proceedings of the 2003 ERAS conference (pp. 144-157). Singapore: Educational Research of Association of Singapore. Gu, Y., Hu, G. W., & Zhang, L. J. (2005). Investigating language learner strategies among lower primary school pupils in Singapore. Language and Education. 19, 281-303. Guerrero, M., & Villamil, O. (2000). Activating the ZPD: Mutual scaffolding in L2 peer revision. The Modern Language Journal, 84(1), 51-68. Gunning, P., & Oxford, R. L. (2014). Children’s learning strategy use and the effects of strategy instruction on success in learning ESL in Canada. System, 43, 82-100. Harris, K. R., Alexander, P., & Graham, S. (2008). Michael Pressley’s contribution to the history and future of strategies research. Educational Psychologist, 43, 86-96. Harris, K. R., Santangelo, T., & Graham, S. (2010). Metacognition and strategies instruction in writing. In H. S. Waters & W. Schneider (Eds.), Metacognition, strategy use, and instruction (pp. 226-256). New York: The Guilford Press.

References

203

Harris, V. (2003). Adapting classroom-based strategy instruction to a distance learning context. TESL-EJ, 7(2), A-1. Hawkins, M. (2004). Researching English language and literacy development in schools. Educational Researcher, 33(3), 14-25. Hickey, D. T., & Granade, J. B. (2004). The influence of sociocultural theory on our theories of engagement and motivation. In D. M. McInerney & S. V. Etten (Eds.), Research on sociocultural influences on motivation and learning: Volume 4 Big theories revisited (pp. 223247). Greenwich, CT: Information Age. Hong-Nam, K., & Leavell, A. (2006). Language learning strategy use of ESL students in an intensive English learning context. System, 34(3), 399-415. Hsiao, T-Y., & Oxford, R. L. (2002). Comparing theories of language learning strategies: A confirmatory factor analysis. The Modern Language Journal, 86(3), 368-383. Hu, G. W. (1999). Explicit metalinguistic knowledge at work: The case of spontaneous written production by formal adult Chinese learners of English. Unpublished doctoral dissertation. Nanyang Technological University, Republic of Singapore. Hu, G. W. (2005). Using peer review with Chinese ESL student writers. Language Teaching Research, 9(3), 321-342. Hu, G.W., & Chen, B. (2007). A protocol-based study of university-level Chinese EFL learners’ writing strategies. English Australia Journal, 23(2), 37-56. Huck, S. W. (2008). Reading statistics and research (5th ed.). Boston: Pearson Education. Humes, A. (1983). Research on composing process. Review of Educational Research, 53(2), 201-216. Ikeda, M., & Takeuchi, O. (2003). Can strategy instruction help EFL learners to improve reading ability? An empirical study. JACET Bulletin, 37, 49-60. Ikeda, M., & Takeuchi, O. (2006). Classifying the differences in learning EFL reading strategies: An analysis of portfolios. System, 34(3), 384398. Johnson, D. M. (1992). Approaches to research in second language learning. New York: Longman. Johnson, K. (1996). Language teaching and skill learning. Cambridge, MA: Blackwell. Johnson, K. (Ed.). (2005). Expertise in second language learning and teaching. Basingstoke: Palgrave Macmillan.

204

References

Jones, S., & Tetroe, J. (1987). Composing in a second language. In A. Matsuhashi (Ed.), Writing in real time: Modeling the production of processes (pp.34-57). Norwood, NJ: Ablex. Harris, K., Graham, S., Mason, L., & Friedlander, B. (2008). Powerful writing strategies for all students. Baltimore, MD: Paul H. Brookes. Kern, R. (2000). Literacy and language teaching. Oxford, UK: Oxford University Press. Kozulin, A. (1994). Cognitive revolution in learning: Piaget and Vygotsky. In J. N. Mangieri & C. C. Block (Eds.), Creating powerful thinking in teachers and students: Diverse perspectives (pp. 269-87). Forth Worth, TX: Harcourt Brace. Krapels, A. R. (1990). An overview of second language writing process research. In B. Kroll (Ed.), Second Language Writing (pp. 37-56). New York: Cambridge University Press. Kuba, A. (2002). Strategies-based instruction in Japanese high schools. Unpublished M.A. Thesis, San Francisco State University, USA. Kucer, S. (2001). Dimensions of literacy: A conceptual base for teaching reading and writing in school settings. Mahwah, NJ: Lawrence Erlbaum Associates. Kucer, S. (2005). Dimensions of literacy: A conceptual base for teaching reading and writing in school settings (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Kucer, S. (2009). Dimensions of literacy: A conceptual base for teaching reading and writing in school settings (3rd ed.). New York: Routledge. Kucer, S., & Silva, C. (2006). Teaching the dimensions of literacy. Mahwah, NJ: Lawrence Erlbaum Associates. Kyoung, R. L., & Oxford, R. L. (2008). Understanding EFL learners’ strategy use and strategy awareness. The Asian EFL Journal, 10, 7-32. Lan, R., & Oxford, R. L. (2003). Language learning strategy profiles of elementary school students in Taiwan. IRAL, 41, 339-379. Lantolf, J. P. (2000). Introducing sociocultural theory. In J. P. Lantolf (Ed.), Sociocultural theory and second language learning (pp.1-26). Oxford, UK: Oxford University Press. Lantolf, J. P. (2006). Sociocultural theory and L2: State of the art. Studies in Second Language Acquisition, 28(1), 67-109. Lantolf, J. P., & Thorne, S. L. (2006). Sociocultural theory and the genesis of second language development. Oxford: Oxford University Press. Lee, C. C. (2002). Strategy and self-regulation instruction as contributors to improving students’ cognitive model in an ESL program. English for Specific Purposes, 21(3), 261-289.

References

205

Leki, I. (1992). Understanding ESL writers: A guide for teachers. Portsmouth, NH: Boynton/Cook. Leki, I. (1995). Coping strategies of ESL students in writing tasks across the curriculum. TESOL Quarterly, 29(2), 235-260. Leow, B. G. (2001). Census of population 2000 statistical release 1: Demographic characteristics. Retrieved from http://www.singstat.gov.sg/pubn/popn/c2000sr1/t1-7.pdf LoCastro, V. (1994). Learning strategies and learning environments. TESOL Quarterly, 28(2), 409-414. Loh, J. M. (2007). Language learning strategies of primary six pupils in a Singapore Primary School. Unpublished M.A. Thesis, National Institute of Education, Nanyang Technological University, Singapore. Luke, A. (1995). When basic skills and information processing just aren't enough: Rethinking reading in new times. Teachers College Record, 97(1), 95-115. Luke, A. (1998). Getting over method: Literacy teaching as work in "New Times". Language Arts, 75(4), 305-313. Luke, A. (2001). Critical literacy in Australia: A matter of context and standpoint. Journal of Adolescent & Adult Literacy, 43(5), 448-461. Macaro, E. (2001). Learning strategies in foreign and second language classrooms. London: Continuum. Macaro, E. (2006). Strategies for language learning and for language use: Revising the theoretical framework. The Modern Language Journal, 90(3), 320-337. Macaro, E. (2007). Do near-beginner learners of French have any writing strategies? Language Learning Journal, 35(1), 23-35. Macaro, E. (2009). Developments in language learner strategies. In W. Li & V. Cook (Eds.), Contemporary Applied Linguistics: Volume 1 Language Teaching and Learning (pp. 10-36). London, UK: Continuum. Macaro, E., & Erler, L. (2008). Raising the achievement of youngbeginner readers of French through strategy instruction. Applied Linguistics, 29(1), 90-119. MacArthur, C. A. (2007). Best practices in teaching evaluation and revision. In S. Graham, C. A. MacArthur, & J. Fitzgerald (Eds.), Best practices in writing instruction (pp. 141-162). New York: The Guilford Press. Mackey, A., & Gass, S. (2005). Second language research: Methodology and design. New Jersey: Lawrence Erlbaum Associates.

206

References

Magogwe, J. M., & Oliver, R. (2007). The relationship between language learning strategies, proficiency, age and self-efficacy beliefs: A study of language learners in Botswana. System, 35(3), 338-352. ManchÓn, R. (2001). Trends in the conceptualizations of second language composing strategies: A critical analysis. International Journal English Studies, 1(2), 47-70. Martin, A. (2008). Enhancing student motivation and engagement: The effects of a multidimensional intervention. Contemporary Educational Psychology, 33(2), 239-269. Mason, L. (2007). Introduction: Bridging the cognitive and sociocultural approaches in research on conceptual change: Is it feasible?. Educational Psychologist, 42(1), 1-7. McCutchen, D. (2006). Cognitive factors in the development of children’s writing. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (pp. 115-130). New York: The Guilford Press. McCutchen, D., Teske, P., & Bankston, C. (2008). Writing and cognition: Implications of the cognitive structure for learning to writing and writing to learn. In C. Bazerman (Ed.), Handbook of research on writing: History, society, school, individual, text (pp. 451-470). New York: Lawrence Erlbaum Associates. McDonough, S. H. (1995). Strategies and skill in learning a foreign language. London, UK: Edward Arnold. McDonough, S. H. (1999). Learner strategies. Language Teaching, 32(1), 1-18. McDonough, S. H. (2005). Training language learning expertise. In K. Johnson (Ed.), Expertise in second language learning and teaching (pp. 150-164). Basingstoke: Palgrave Macmillan. McKay, S. (2006). Researching second language classrooms. Mahwah, NJ: Lawrence Erlbaum. McLaughlin, B. (1987). Theories of second-language learning. London, UK: Edward Arnold. McLaughlin, B. (1990). Restructuring. Applied Linguistics, 11, 113-128. McLaughlin, B., & Heredia, R. (1996). Information processing approaches to research on second language acquisition and use. In R. Ritchie & T. Bhatia (Eds.), A handbook of second language acquisition (pp. 213-28). San Diego: Academic Press. Min, H. T. (2006). The effects of trained peer review on EFL students’ revision types and writing quality. Journal of Second Language Writing, 15(2), 118-141.

References

207

Mizumoto, A., & Takeuchi, O. (2009). Examining the effectiveness of explicit instruction of vocabulary learning strategies with Japanese EFL university students. Language Teaching Research, 13(4), 425-449. Naiman, N., Fröhlich, M., Stern, H. H., & Todesco, A. (1978). The good language learner. Toronto: Ontario Institute for Studies in Education. Nesbit, J. C., & Hadwin, A. F. (2006). Methodological issues in educational psychology. In P. A. Alexander & P. H. Winne (Eds.), Handbook of educational psychology (2nd ed., pp. 825-847). Mahwah, NJ: Lawrence Erlbaum Associates. Nisbet, J., & Shucksmith, J. (1986). Learning Strategies. London: Routledge & Kegan Paul. Nisbet, D. L., Tindall, E. R., & Arroyo, A. A. (2005). Language learning strategies and English proficiency of Chinese university students. Foreign Language Annals, 38, 100-107. Norušis, M. J. (2009). SPSS 17.0 Guide to Data Analysis. Chicago: Prentice Hall Inc. Norton, B., & Toohey, K. (2001). Changing perspectives on good language learners. TESOL Quarterly, 35(2), 307-322. Nunan, D. (1997). Strategy training in the language classroom: An empirical investigation. RELC Journal, 28(2), 56-81. O'Malley, J. M., & Chamot, A. U. (1990). Learning strategies in second language acquisition. Cambridge: Cambridge University Press. O'Malley, J. M., Chamot, A. U., & Küpper, L. (1989). Listening comprehension strategies in second language acquisition. Applied Linguistics, 10(4), 418-435. O'Malley, J. M., Chamot, A. U., Stewner-Manzanares, G., Küpper, L., & Russo, R. P. (1985a). Learning strategies used by beginning and intermediate ESL students. Language Learning, 35(1), 21-46. O'Malley, J. M., Chamot, A. U., Stewner-Manzanares, G., Küpper, L., & Russo, R. P. (1985b). Learning strategy applications with students of English as a second language. TESOL Quarterly, 19(3), 557-584. O'Malley, J. M., Chamot, A. U., Walker, C. (1987). Some applications of cognitive theory to second language acquisition. Studies in Second Language Acquisition, 9(3), 287-306. Oxford, R. L. (1989). Use of language learning strategies: A synthesis of studies with implications for strategy training. System, 17(2), 235-247. Oxford, R. L. (1990). Language learning strategies: What every teacher should know. New York: Newbury House. Oxford, R. L., Cho, Y., Leung, S., & Kim, H-J (2004). Effect of the presence and difficulty of task on strategy use: An exploratory study. International Review of Applied Linguistics, 42, 1-47.

208

References

Oxford, R. L., & Crookall, D. (1989). Research on language learning strategies: Methods, findings, and instructional implications. The Modern Language Journal, 73(4), 404-419. Oxford, R. L., & Ehrman, M. (1995). Adults’ language learning strategies in an intensive foreign language program in the United States. System, 23(3), 359-386. Oxford, R. L., & Schramm, K. (2008). Bridging the gap between psychological and sociocultural perspectives on L2 learner strategies. In A. D. Cohen & E. Macaro (Eds.), Language Learner Strategies: 30 years of research and practice (pp. 47-68). Oxford, UK: Oxford University Press. Padron, Y., & Waxman, H. (1988). The effect of ESL students’ perceptions of their cognitive strategies on reading achievement. TESOL Quarterly, 22(1), 146-150. Perl, S. (1979). The composing process of unskilled college writers. Research in the Teaching of English, 3, 317-336. Petriü, B., & Czárl, B. (2003). Validating a writing strategy questionnaire. System, 31, 187-215. Politzer, R. L. (1983). Research notes: An exploratory study of selfreported language learning behaviors and their relation to achievement. Studies in Second Language Acquisition, 6, 54-68. Politzer, R. L., & McGroarty, M. (1985). An exploratory study of learning behaviors and their relationship to gains in linguistic and communicative competence. TESOL Quarterly, 19, 103-123. Pressley, M., & Harris, K. R. (2006). Cognitive strategies instruction: From basic research to classroom instruction. In P. A. Alexander & P. Winne (Eds.), Handbook of educational psychology (2nd ed., pp. 265286). New York: Macmillan. Pressley, M., Mohan, L., Fingeret, L., Reffitt, K., & Raphael-Bogaert, L. (2007). Writing instruction in engaging and effective elementary settings. In S. Graham, C. A. MacArthur, & J. Fitzgerald (Eds.), Best practices in writing instruction (pp. 13-27). New York: The Guilford Press. Pritchard, R. J., & Honeycutt, R. L. (2007). Best practices in implementing a process approach to teaching writing. In S. Graham, C. A. MacArthur, & J. Fitzgerald (Eds.), Best practices in writing instruction (pp. 28-49). New York: The Guilford Press. Purdie, N., & Oliver, R. (1999). Language learning strategies used by bilingual school-aged children. System, 27, 375-388. Rabinowitz, M., & Chi, M. T. H. (1987). An interactive model of strategic processing. In S. J. Cesi (Ed.), Handbook of cognitive, social,

References

209

neuropsychological aspects of learning disabilities (Vol. 21, pp. 83102). New York: Academic Press. Rao, Z. (2002). Bridging the gap between teaching and learning styles in East Asian context. TESOL Journal, 11, 5-11. Rao, Z., Gu, Y., Zhang, L. J., & Hu, G. W. (2007). Reading strategies and approaches to learning of bilingual primary school pupils. Language Awareness, 16(4), 243-262. Raupach, M. (1987). Procedural learning in advanced learners of a foreign language. In J. Coleman & R. Towell (Eds.), The advanced language learner (pp. 123-155). London, UK: AFLS/SUFLRA/CILT. Roca de Larios, J., ManchÓn, R., Murphy, L., & Marín, J. (2008). The foreign language writer’s strategic behavior in the allocation of time to writing processes. Journal of Second Language Writing, 17(1), 30-47. Rowe, D. W. (2008). Development of writing abilities in childhood. In C. Bazerman (Ed.), Handbook of research on writing: History, society, school, individual, text (pp. 401-420). New York: Lawrence Erlbaum Associates. Rubin, J. (1975). What the good language learner can tell us. TESOL Quarterly, 9(1), 41-51. Rubin, J. (1981). Study of cognitive processes in second language learning. Applied Linguistics, 11(2), 117-131. Rubin, J. (1987). Learner strategies: Theoretical assumptions, research history, and typology. In A. Wenden & J. Rubin (Eds.), Learner strategies in language learning (pp. 15-30). Englewood Cliffs, N.J.: Prentice-Hall. Rubin, J. (2005). The expert language learner: A review of good language learner studies and learner strategies. In K. Johnson (Ed.), Expertise in second language learning and teaching (pp. 37-63). Basingstoke: Palgrave Macmillan. Sasaki, M. (2000). Toward an empirical model of EFL writing processes: an exploratory study. Journal of Second Language Writing, 9(3), 259291. Schmeck, R. R. (1988). An introduction to strategies and styles of learning. In R. R. Schmeck (Ed.), Learning strategies and learning styles (pp. 319). New York: Plenum Press. Seng, A. S. H., Pou, L. K. H., & Tan. O. S. (Eds.). (2003). Mediated learning experience with children: Applications across contexts. Singapore: McGraw-Hill. Sengupta, S. (2000). An investigation of the effects of revision strategy instruction on L2 secondary school learners. System, 28(1), 97-113.

210

References

Shapira, A., & Hertz-Lazarowitz, R. (2005). Children’s writing strategies: Profiles of writers. In G. Rijlaarsdam, H. van den Bergh, & M. Couzijn (Eds.), Effective learning and teaching of writing: A handbook of writing in education (2nd ed., pp. 574-586). New York: Kluwer Academic. Shuell, T. J. (1986). Cognitive conceptions of learning. Review of Educational Research, 56, 411-436. Silva, T. (1990). Second language composition instruction: developments, issues, and directions in ESL. In B. Kroll (Ed.), Second language writing (pp. 11-21). New York: Cambridge University Press. Singapore Department of Statistics. (2011). Census of population 2010: Statistical release 1 – demographic characteristics, education, language and religion. Singapore: Author. Skehan, P. (1989). Individual differences in second language learning. London, UK: Edward Arnold. Smith, F. (1994). Understanding reading (5th ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Smith, F. (2004). Understanding reading (6th ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Stern, H. H. (1975). What can we learn from the good language learner? Canadian Modern Language Review, 31(4), 304-318. Stern, H. H. (1983). Fundamental concepts of language teaching. Oxford: Oxford University Press. Sugeng, B. (1997). A learning strategy profile of Indonesian elementary school students. RELC Journal, 28(2), 82-106. Takeuchi, O. (2003). What can we learn from good language learners?: A qualitative study in the Japanese foreign language context. System, 31, 313-432. Tseng, W. T., Dörnyei, Z., & Schmitt, N. (2006). A new approach to assessing strategic learning: The case of self-regulation in vocabulary acquisition. Applied Linguistics, 27(1), 78-102. Tzuriel, D. (2000). Developmental perspectives of mediated learning experience theory. In A. Kozulin & Y. Rand (Eds.). Experience of mediated learning: An impact of Feuerstein’s theory in education and psychology (pp. 217-39). Oxford, UK: Pergamon. Urzua, C. (1987). “You stopped too soon”: Second language children composing and revising. TESOL Quarterly, 21(2), 279-303. Vandergrift, L. (1997). The comprehension strategies of language (French) learners: A descriptive study. Foreign Language Annals, 30, 387-409.

References

211

Vandergrift, L. (2003a). From prediction through reflection: Guiding students through the process of L2 listening. Canadian Modern Language Review, 59(3), 425-440. Vandergrift, L. (2003b). Orchestrating strategy use: Toward a model of the skilled second language listener. Language Learning, 53, 463-496. Victori, M. (1999). An analysis of writing knowledge in EFL composing: a case study of two effective and two less effective writers. System, 27(4), 537-555. Vygotsky, L. S. (1962). Thought and language. Cambridge, MA: MIT Press. Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: Harvard University Press. Vygotsky, L. S. (1981). The development of higher forms of attention in childhood. In J. V. Wertsch (Ed.), The concept of activity in Soviet psychology (189-240). Armonk, NY: Sharpe. Vygotsky, L. S. (1986). The genetic roots of thought and speech. In A. Kozulin (Trans. & Ed.), Thought and language. Cambridge, MA: MIT Press. Wallace, C. (2005). Reading and expertise. In K. Johnson (Ed.), Expertise in second language learning and teaching (pp. 85-103). Basingstoke: Palgrave Macmillan. Walters, S. (2007). Researching Bangladeshi pupils’ strategies for learning to read in (UK) primary school settings. Language Learning Journal, 35(1), 51-64. Wang, W. Y., & Wen, Q. F. (2002). L1 use in the composing process: An exploratory study of 16 Chinese EFL writers. Journal of Second Language Writing, 11(3), 225-246. Weigle, S. C. (2002). Assessing writing. Cambridge, UK: Cambridge University Press. Weigle, S. C. (2005). Second language writing expertise. In K. Johnson (Ed.), Expertise in second language learning and teaching (pp. 128149). Basingstoke: Palgrave Macmillan. Weigle, S. C. (2007). Teaching writing teachers about assessement. Journal of Second Language Writing, 16(3), 194-209. Weinstein, C. E., & Mayer, R. E. (1986). The teaching of learning strategies. In M. C. Wittrock (Ed.), Handbook of research on teaching (3rd ed., pp. 315-327). New York: Macmillan. Weinstein, C. E., Husman, J., & Dierking, D. R. (2000). Self-regulatin interventions with a focus on strategies. In M. Boekaerts, P. R. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 727-747). San Diego, CA: Academic Press.

212

References

Wenden, A. (1986). Incorporating learner training in the classroom. System, 14, 315-325. Wenden, A. (1987a). Conceptual background and utility. In A. Wenden & J. Rubin (Eds.), Learner strategies in language learning (pp. 3-13). London, UK: Prentice Hall. Wenden, A. (1987b). Incorporating learner training in the classroom. In A. Wenden & J. Rubin (Eds.), Learner strategies in language learning (pp. 3-13). London, UK: Prentice Hall. Wenden, A. (1991). Learner strategies for learner autonomy. New York: Prentice Hall. Wenden, A., & Rubin, J. (Eds.). (1987). Learner strategies in language learning. London, UK: Prentice Hall. Wertsch, J. V. (Ed.). (1981). The concept of activity in Soviet psychology. Armonk, NY: Sharpe. Wharton, G. (2000). Language learning strategy use of bilingual foreign language learners in Singapore. Language Learning, 50(2), 203-243. Wong-Fillmore, L. (1976). The second time around: Cognitive and social strategies in second language acquisition. Unpublished PhD thesis, Stanford University, US. Wong-Fillmore, L. (1979). Individual differences in second language acquisition. In C. J. Fillmore, W. S-Y. Wang, & D. Kempler (Eds.), Individual differences in language ability and language behavior. New York: Academic Press. Woodrom, L. (2005). The challenge of measuring language learning strategies. Foreign Language Annals, 38, 90-99. Yang, M. N. (2007). Language learning strategies for junior college students in Taiwan: Investigating Ethnicity and Proficiency. The Asian EFL Journal, 9, 35-57. Zamel, V. (1983). The composing processes of advanced ESL students: Six case studies. TESOL Quarterly, 17, 165-187. Zhang, L. J. (2003). Research into Chinese EFL learners strategies: Methods, findings and instructional issues. RELC Journal, 34(3), 284322. Zhang, L. J. (2008). Constructive pedagogy in strategic reading instruction: Exploring pathways to learner development in the English as a second language (ESL) classroom. Instructional Science, 36, 89-116. Zhang, L. J. (2010). A dynamic metacognitive systems account of Chinese university students’s knowledge about EFL reading. TESOL Quaterly, 44(2), 320-353. Zhang, L. J., Gu, P., & Hu, G. (2008). A cognitive perspective on Singaporean primary school pupils’ use of reading strategies in

References

213

learning to read. British Journal of Educational Psychology, 78(2), 245-271. Zhang, L. J., & Xiao, Y. H. (2006). Language learning strategies, motivation, and EFL proficiency: A study of Chinese tertiary-level non-English majors. Asian Englishes, 9, 20-47. Zito, J., Adkins, M., Gavins, M., Harris, K., & Graham, S. (2007). Selfregulated strategy development: Relationship to the social-cognitive perspective and the development of self-regulation. Reading & Writing Instruction, 23, 77-95. Zuengler, J., & Miller, E. R. (2006). Cognitive and sociocultural perspectives: Two parallel SLA worlds?. TESOL Quarterly, 40(1), 3558.

APPENDIX A TEACHER’S JOURNAL LOG

Class ID:

Name:

Lesson ID:

No of Students:

Date:

What do you think of this lesson in terms of the topic, time frame, requirements, content, etc.?

Appendix A

Do you have any suggestions to improve the lesson plan?

Please describe and assess how students have reacted to this lesson.

215

216

Appendix A

Please feel free to let us know your other comments.

APPENDIX B RESEARCHER’S OBSERVATION RECORD LOG

Lesson ID:

Researcher’s Name:

Date:

Class ID:

No of Students:

Teacher’s Name:

School:

Activities as they occurred.

Each event must be accompanied by the time that it happened.

218

Appendix B

Reflections on teacher’s delivery of the SBI lesson

Students’ participation/response in this lesson.

Appendix B

Do you have any other comments on the lesson.

219