比較台灣大學生對中文及英文閱讀信念與閱讀策略之研究

全文

(1)國立台灣師範大學英語學系博士論. 文. Doctoral Dissertation Graduate Institute of English National Taiwan Normal University. 比較臺灣大學生對中文及英文閱讀信念與閱讀策略之研究. Chinese EFL College Students’ Beliefs about Reading and Reading Strategies in L1 and L2. 指導教授：程玉秀博士、陳秋蘭博士 Advisors: Dr. Yuh-Show Cheng and Dr. Chiou-Lan Chern 研究生：許麗媛. 中華民國九十八年六月 June, 2009. 1.

(2) 中文摘要本研究主要的目的是在比較學生的中文及英文閱讀信念與閱讀策略，及探究兩者之間的關聯。共有 432 位臺灣的大學生參與此研究，本研究所探討的閱讀信念以兩個閱讀信念模式為主軸─情意建構模式及知識傳遞模式。研究者提出四個研究問題，第一個問題試圖比較及了解學生對中文及英文的閱讀信念，第二個問題則是比較及了解學生對中文及英文的閱讀策略，第三個問題企圖探討中文及英文閱讀信念和策略兩者之間的關聯。針對以上的三個問題，資料蒐集採取問卷調查的方式。至於第四個研究問題，研究者試圖針對不同閱讀信念模式的學生，觀察他們閱讀策略實際運用的情況，來進一步了解中文英文閱讀信念和閱讀策略之間的關聯，針對學生的實際閱讀策略運用，研究者是採用放聲思考的方法來蒐集資料。本研究的主要發現為：(一)整體來說，學生在中文閱讀信念量表的得分顯著高於英文量表。並且學生對中文及英文抱有不同的閱讀信念模式，在閱讀中文的時候，學生普遍抱持較強的情意建構信念模式；然而在閱讀英文的時候，學生只有對情意建構模式中的個人解讀部份之得分高於知識傳遞模式，至於情意建構模式中的另一類別─情感的投入，在學生閱讀英文時，則和知識傳遞模式的得分無顯著差別。(二)閱讀策略量表的分析結果顯示，學生的中文及英文閱讀策略之間有明顯的相似以及差異之處。就相似處而言，學生不管在閱讀中文或英文時，都普遍喜歡使用的閱讀策略有三個：運用圖表來幫助理解，將內容與已知的相連結，以及來回重讀。就差異性來說，學生在讀中文時會使用較多的閱讀策略，並且會用較高比例的統整型策略；但是在讀英文時，則使用較少策略，並且對非統整型策略的運用比例較高。(三)研究結果顯示學生中文及英文的閱讀信念和閱讀策略兩者之間有顯著的正相關；此外雖然研究中發現，情意建構模式信念較強的學生會使用較多的策略，而知識傳遞模式信念較強的學生則使用較少的策略，但是各組學生之間的差異未達統計上顯著之標準。(四)放聲思考的資料分析結果顯示，不管在閱讀中英文時，有著不同閱讀信念模式的學生，會展現出不同的策略運用型態。情意建構模式信念較強的學生容易運用較多高層次的閱讀策略，而相反的，知識傳遞模式信念較強的學生則容易使用較多低層次的閱讀策略。然而研究發現此一差異性，在學生閱讀中文時比閱讀英文時明顯，因為學生在閱讀英文時，不管那一類型信念模式的學生，都過度局限於使用少數幾種區域型的閱讀策略。. i.

(3) ABSTRACT The purpose of this study was to examine learners’ beliefs about reading and reading strategies in L1 and L2. Four hundred and thirty-two Taiwanese college students participated in this study. The primary concern was to investigate whether learners’ beliefs about reading relate to their use of reading strategy in L1 (Chinese) and L2 (English) when reading for academic purposes. Readers’ beliefs were investigated in terms of the two implicit models of reading, the transaction and transmission beliefs. Four major research questions were addressed in the study. The first research question concerned students’ beliefs about reading in L1 and L2. The second inquiry attempted to compare students’ reported use of reading strategies between languages. The third question intended to investigate whether learners’ beliefs about reading were related to their reported use of reading strategies in L1 and L2. To address these three questions, four measurement questionnaires were developed in this study to collect data. Lastly, since the fourth research question aimed to examine whether students with distinctive beliefs about reading in the two languages show any differences in how they process texts, the think-aloud method was used to collect information about students’ actual use of reading strategies. The findings of the study were: (1) Students showed somewhat different reading belief patterns between L1 and L2. More specifically, when reading in L1, students tended to hold stronger transaction than transmission beliefs. On the other hand, when reading in L2, students were found to show differential strengths of two kinds of transaction beliefs identified in the study, with stronger endorsement to Transaction I beliefs (i.e. Reader Interpretation) than to Transaction II beliefs (i.e. Reader Engagement); no difference was found, however, between students’. ii.

(4) Transaction II and Transmission beliefs. Furthermore, in general, students were found to hold stronger beliefs about reading in L1 than in L2, either transaction or transmission beliefs. (2) Both similarities and differences were found in students’ reported use of reading strategies between L1 and L2. Concerning the similarities between the languages, three strategies including using tables and pictures, linking content with what one already knows, and re-reading were found to be frequently used by students no matter whether they read in L1 or L2. As for the differences, students reportedly applied significantly more reading strategies in L1 than in L2. Moreover, while students reportedly used a significantly higher amount of general strategies when reading in L1, they tended to apply a significantly larger proportion of non-general strategies when reading in L2. (3) Students’ beliefs about reading and their reported use of strategies were found to be positively correlated in either L1 or L2. More specifically, both transaction and transmission beliefs were significantly correlated with the two reading strategy categories (i.e. general versus non-general). Regarding the reported uses of strategies by readers with distinctive belief types, even though there was a tendency for readers with strong transaction beliefs to report more use of reading strategies than those with strong transmission beliefs in either L1 or L2, the differences among groups were not statistically significant. (4) The qualitative results of the study have revealed that students with distinctive belief types in both languages seemed to exhibit distinctive patterns of strategy use. There was a tendency for readers with high transaction beliefs to process the text at a higher level and those with high transmission beliefs at a lower level. The differences, however, were more apparent in L1 than in L2. When reading in L2, although readers with distinctive belief types showed different strategy patterns, in general, they all tended to heavily rely on several local strategies.. iii.

(5) ACKNOWLEDGMENTS. Many people have contributed to this study. First of all, I would like to express my deepest gratitude to my advisors, Dr. Yuh-Show Cheng and Dr. Chiou-Lan Chern, for their guidance and encouragement during my entire process of writing this dissertation. Their valuable suggestions helped me make improvements in this work. I am also very grateful to my committee members, Dr. Hsi-Chin Chu, Dr. Chin-Kuei Cheng, and Dr. Nae-Dong Yang, for their interest in this study and for their great comments. I would also like to give my special thanks to Ms. Hsiu-Chuan Chen for not only helping me with the data coding but also giving me lots of emotional support. My thanks also go to Ms. Shu-Li Lai for her friendship and great assistance. In addition, my appreciation also goes to Dr. Jane Zuengler for providing me with the wonderful opportunity to visit the University of Wisconsin-Madison and compose the major parts of this work there. I must also express my gratitude to the participants of this study. Without their cooperation, the study could not have been conducted. Finally, I would like to dedicate this dissertation to my husband, Dr. Wei-Neng Lin, my daughter, Sammie, and my Mom. Their love and support throughout my many years of studying in the doctoral program have given me the strength needed to complete this dissertation.. iv.

(6) TABLE OF CONTENTS. LIST OF TABLES………………………………………………………………. viii. CHAPTER ONE: INTRODUCTION…………………………………………... Background………………………………………………………………... Rationale…………………………………………………………………... Definitions of Beliefs and Strategies………………………………………. Purpose of the Study………………………………………………………. Significance of the Study…………………………………………………... 1 1 3 6 7 8. CHPATER TWO: LITERATURE REVIEW…………………………………… Major L1 Reading Theories……………………………………………….. Three Often-Cited Cognitive Models of Reading……………………. Rosenblatt’s Transactional Model of Reading……………………….. Two Affective Models of Reading…………………………………… Reading Beliefs in L1………….………………………………………….. Reading Strategies in L1…………………………………………………... Learner Beliefs in SLA……………………………………………………. Reading Beliefs in L2………..……………………………………………. Reading Strategies in L2............................................................................... Studies Using Think-Aloud……………………..……………………. Studies Using Questionnaires……………………………………….... 10 10 10 13 14 16 18 21 23 26 26 36. CHAPTER THREE: METHODOLOGY………………………………………. Participants………………………………………………………………… Instruments………………………………………………………………… The Reader Belief Questionnaire (RBQ)…………………………….. The Reading Strategy Questionnaire (RSQ)…………………………. The Reading Passages………………………………………………... Data Collection Procedures………………………………………………... Data Analysis……………………………………………………………… Analysis of the Questionnaires………………………………………. Analysis of the Think-Aloud Protocols………………………………. 40 40 42 43 52 62 65 66 66 67. v.

(7) CHAPTER FOUR: RESULTS………………………………………………….. Background Information of the Participants………………………………. Readers’ Beliefs in L1 and L2……………………………………............... Reported Use of Reading Strategies in L1 and L2………………………… The Relationship Between the Readers’ Beliefs and Reported Use of Reading Strategies in L1 and L2………………………… The Relationship Between the Readers’ Beliefs and Actual Use of Reading Strategies in L1 and L2……………………………. 70 70 70 73. CHAPTER FIVE: DISCUSSION AND CONCLUSION………………………. Overview of the Study…………………………………………………….. Discussion…………………………………………………………………. Conclusions………………………………………………………………... Implications for Future Research………………………………………...... Implications for Instruction………………………………………………... Limitations…………………………………………………………………. 137 137 138 151 153 155 157. REFERENCES………………………………………………………………….. 158. Appendix A: The Background Questionnaire ………………………………….. Appendix B: The Preliminary Chinese Reader Belief Questionnaire (RBQ)……………………………………… Appendix C: The Preliminary English Reader Belief Questionnaire (RBQ)……………………………………… Appendix D: The Finalized 9-Item Reader Belief Questionnaire (RBQ)………. Appendix E: The CFA Results for the Finalized Chinese RBQ…..…………….. Appendix F: The CFA Results for the Finalized English RBQ…………………. Appendix G: The Preliminary Chinese Reading Strategy Questionnaire (RSQ)……………………………………. Appendix H: The Preliminary English Reading Strategy Questionnaire (RSQ)……………………………………. Appendix I: The Finalized 29-Item Reading Strategy Questionnaire (RSQ)…... Appendix J: The Results of CFA for the Finalized Chinese RSQ………………. Appendix K: The Results of CFA for the Finalized English RSQ……………… Appendix L: The English Reading Passages…………………..……………….. Appendix M: The Chinese Reading Passages……………..……………………. Appendix N: The Strategy Classification Scheme………………......................... 168. vi. 82 88. 169 171 173 174 175 176 181 186 188 189 190 191 192.

(8) Appendix O: The Number of Occurrence for Individual Strategy in L1 Reading……………………………………………… Appendix P: The Number of Occurrence for Individual Strategy in L2 Reading………………………………………………. vii. 195 196.

(9) LIST OF TABLES. Table 1.. Rotated Factor Patterns of the Preliminary 12-Item Chinese and English Reader Belief Questionnaires (RBQs)………… Table 2. Rotated Factor Patterns of the Preliminary 9-Item Chinese and English Reader Belief Questionnaires (RBQs)………… Table 3. Rotated Factor Patterns of the Preliminary 43-Item Chinese and English RSQs…………………………………………... Table 4. Rotated Factor Patterns of the Preliminary 29-Item Chinese and English RSQs…………………………………………... Table 5. Mean Scores and Standard Deviations for the Reading Belief Types in L1 and L2………………………………….. Table 6. Summary of T Tests for the Belief Types Within and Across Languages………………………………………... Table 7. Means of Reported Reading Strategy Use in L1 In Descending Order…………………………………………………. Table 8. Means of Reported Reading Strategy Use in L2 In Descending Order…………………………………………………. Table 9. The Five Most and Least Frequently Reported Use of Reading Strategies in L1 and L2……………………………………... Table 10. Mean Scores and Standard Deviations for the Overall Strategy Scales and Subscales in L1 and L2………………… Table 11. Summary of T Tests for the Reported Strategy Use Within and Across Languages………………………………………... Table 12. Summary of Pearson Correlations Between Belief And Reported Strategy Use in L1……………………………………. Table 13. Summary of Pearson Correlations Between Belief And Reported Strategy Use in L2……………………………………. Table 14. Mean Scores and Standard Deviations of Reported Use of Reading Strategies in L1 by the Four Types of Readers……………... Table 15. Mean Scores and Standard Deviations of Reported Use of Reading Strategies in L2 by the Four Types of Readers……………... Table 16. Frequencies of Reading Strategy Used Across the Two Languages………………………………………………………. Table 17. The Five Most Frequently Used Stategies Across the Two Languages……………………………………………………….. viii. 45 48 53 58 72 73 75 77 79 81 82 84 85 87 88 90 91.

(10) Table 18. Frequencies of Reading Strategies Used by the Three Reader Types and Across the Two Languages……………………….. Table 19. The Five Most Frequently Used Strategies by Type – A (TA-TA) Readers in L1 and L2…………………………….. Table 20. Hannah’s Use of Reading Strategies in L1 and L2…………………... Table 21. The Five Most Frequently Used Strategies by Type B (TA-TM) Readers in L1 and L2……………………………... Table 22. Jane’s Use of Reading Strategies in L1 and L2………………………. Table 23. The Five Most Frequently Used Strategies by Type C (TM-TM) Readers in L1 and L2…………………………….. Table 24. Tina’s Use of Reading Strategies in L1 and L2……………………….. ix. 93 95 104 113 120 128 131.

(11) CHAPTER ONE. INTRODUCTION. Reading in a second language (L2) has long been the interest of many teaching professionals and researchers. L2 reading ability, particularly with English as the L2, is considered to be fundamental as English continues to spread, not only as a global language but also as the language of science and technology as well as other advanced disciplines. A multitude of studies have been conducted in the past few decades in seeking for ways to promote effective reading in a second language. More recently, as Koda (2005) pointed out, the research focus progressively shifted from the mere description of what L2 readers could or could not do to the explanation of the contributory factors behind those behaviors. In addition, new research trends in second language acquisition (SLA) have also begun to recognize the importance of the affective dimension within individual learners. The present study, in joining the current conceptual trends, attempted to investigate one of the affective factors in reading. It aimed to examine Taiwanese students’ beliefs about reading and their use of reading strategies in both their first language (L1), Chinese, and L2, English. The study also explored the possible links between beliefs about reading and the use of reading strategies in both contexts. Background As an English instructor in Taiwan, I have long been interested in exploring the theories in teaching reading and learning to read. In particular, I am concerned with the possible factors that contribute to Taiwanese learners’ achievement of reading in English as a foreign language (EFL). Based on my own teaching experiences as well as informed by the findings from the field of reading research, I have identified several factors for students’ success or failure in L2 reading, especially regarding their. -1-.

(12) learning from texts in the academic setting. These factors range from the characteristics of individual learner such as L2 proficiency, L1 reading ability, L2 vocabulary size, language aptitude, and motivation to the external issues such as reading instruction and language policy. The various investigations into these variables have constituted a considerable body of research on L2 literacy. In more recent years, the notion of “learner beliefs” has received much attention in the fields of general education, L1 reading, and language teaching. Researchers from the general education domain suggested that students’ success in school cannot be explained by cognitive factors alone (Zimmerman, 1990). Moreover, learners’ beliefs about the nature of knowledge and learning may influence how they approach learning (Schommer, 1994). In the area of L1 reading, researchers discovered that beliefs about reading are related to understanding and engagement during reading (Schraw & Bruning, 1999), have distinct effects on comprehension (Schommer, 1990), and are part of the meaning-making process (Wade, Thompson, & Watkings, 1994). In addition, studies from the field of second language acquisition have shown that learners’ preconceived beliefs about language learning have the potential to influence the learning processes and their choice of learning strategies (Horwitz, 1987, 1988, 1999; Wenden, 1987; Yang, 1999). Finally, a number of studies on L2 reading also pointed out that students’ beliefs about reading are closely related to their use of reading strategies (Devine, 1988; Kamhi-Stein, 2003; Yang, 2004). Motivated by the theories of language learning and literacy development as well as my long-term interest in reading instruction, the present study intended to investigate Taiwanese EFL learners’ beliefs about reading and the relations between beliefs and use of strategies. It also addressed the issue from a cross-linguistic approach by examining L1 and L2 reading processes in tandem. This is largely influenced by the current research development in L2 reading which emphasizes the -2-.

(13) cross-linguistic comparison of reading processes (Davis & Bistodeau, 1993; Horiba, 1996; Stevenson, Schoonen, & de Glopper, 2007; van Gelderen et al., 2004), the impact of L1 processing experiences on L2 reading (Hamada, 2005; Koda, 2005), and the contribution of L1 reading attitude to L2 reading attitude (Yamashita, 2004; 2007). Koda (2005) specifically emphasized the benefits of cross-linguistic analyses, arguing that they “can illuminate the subtle ways in which L1 and L2 experiences meld and interface during L2 reading development” (p.9). The remainder of this section explains the rationale of the study in more details by discussing major antecedents to the present inquiry. Rationale Three strands of research contributed to the current investigation on the relationships between learners’ beliefs about reading and their use of reading strategies. The first one was the series of studies conducted to look at second/foreign language learners’ beliefs about language learning (Horwitz, 1987, 1988, 1999; Wenden, 1987), which demonstrated the potential influence of learner’ beliefs on learning experiences and actions they take as well as their links to various aspects of learning. Influenced by the work of Horwitz, many researchers have examined students’ beliefs about language learning and their connections to different learning variables (Cotterall, 1999; Huang, 2003; Mantle-Bromley, 1995; Peacock, 1999; Sakui & Gaies, 1999; Yang, 1999). In particular, research has found that language learners’ beliefs about learning English are strongly related to their use of strategies (Yang, 1999). Secondly, this research was also greatly inspired by a number of studies on L2 readers’ beliefs (Devine, 1988; Kamhi-Stein, 2003; Yang, 2004), many of which used in-depth qualitative analyses to examine ESL students’ beliefs about reading. Their findings suggested that there was a close relationship between learners’ beliefs about -3-.

(14) reading in English and their use of reading strategies. For instance, Devine’s (1988) study showed that ESL learners’ internalized models of reading, whether it’s the sound-, word-, or meaning-centered model, might influence the ways in which they approach a text and their comprehension of the text. Kamhi-Stein’s (2003) investigation also yielded a similar conclusion that readers who view reading as a process of meaning construction tended to be more multi-strategic, while those who view reading as primarily word-centered tended to be logo-centric during reading. The final line of research that directly informed the present investigation was the work of Schraw and Bruning (1996; 1999), who studied L1 readers of English and showed that readers’ implicit models of reading can affect motivation to read and reading engagement. They distinguished between two different implicit models, transmission and transaction models. The transmission model is characterized by the beliefs that meaning is transmitted from the author and/or text and hence, independent of the reader, while the transaction model is based on the assumption that meaning is constructed by a transaction between the reader, author, and text. The transmission model represents the more traditional view of reading in which the reader is often viewed as a passive receiver whose main goal is to extract meaning from the text or intended by the author (Shraw & Bruning, 1999). Their studies found that a strong commitment to the transaction model increased motivation to read by strengthening the degree to which readers valued the meaning-construction process and by increasing the type and number of deeper processing strategies. Schraw and Bruning (1996; 1999) argued that adult readers possess multiple belief systems that shape the ways they read. Although the results of Schraw and Bruning’s studies offered valuable insights about the role of readers’ beliefs, two areas of their work deserved further exploration. First, although their studies showed that L1 readers’ transaction beliefs about reading connected to higher reader engagement, it is not clear that whether the -4-.

(15) same argument can be extended to L2 readers. Second, their studies only examined the effect of beliefs on the products of reading; namely, they only compared the response essays produced by readers with different belief systems. They did not provide evidence for whether and how the endorsement of certain beliefs might have an influence on one’s reading processes. It is therefore particularly valuable for the present study to investigate the possibility of how one’s beliefs about reading might have an effect on the way they approach a text, i.e. their use of reading strategies. From this brief review of literature, we can conclude that although numerous studies have been conducted in the field of SLA regarding learners’ beliefs about language learning, little research has looked into the issue of L2 readers’ beliefs and their relations to reading processes (Devine, 1988; Kamhi-Stein, 2003; Yang, 2004). In addition, the studies cited regarding L2 learners’ beliefs about reading were mostly carried out in an ESL context in which the subjects mainly consisted of ESL students studying at the U.S. universities. Despite the valuable insights gained from these studies, the application of their findings to an EFL learning context such as Taiwan should be considered limited. Bernhardt (1991) emphasized that “the processing of text can be viewed only within a unique cultural context (p.10).” Any assertions made about readers’ beliefs, therefore, should consider the socio-cultural factors such as the educational context, purposes, as well as L1 literacy instruction. This line of research is, therefore, worth pursuing in an EFL context such as Taiwan where students have already had abundant literary experiences in their first language, Mandarin Chinese, and English is mostly studied as a required course and hardly used outside of English classrooms. Lastly, Schraw and Bruning’s (1996; 1999) studies are particularly influential to the current research in that their distinction of readers’ implicit models of reading, transmission and transaction, offers an alternative perspective and framework for the -5-.

(16) investigation of readers’ beliefs. Although Devine (1988), as previously cited, also proposed the internalized models of reading (i.e. the sound-, word-, and meaning-centered models), it was considered less ideal than Schraw and Bruning’s (1996; 1999) framework simply because Devine’s internalized models were very text-oriented or text-based in nature. They failed to capture the current theoretical perspective that sees readers as having a more active role in the construction of meaning from text. In light of Schraw and Bruning’s (1996; 1999) work, the present study also defined readers’ beliefs in terms of readers’ implicit models of the reading process. The next section provides a more detailed account of the scope of beliefs under investigation and definitions of reading strategies. Definitions of Beliefs and Strategies Readers’ Beliefs Since the term “belief” has been widely used in many areas of research and often been associated with various kinds of concepts and meanings, it is necessary to clarify the definition of it in the study. Following Schraw and Bruning’s (1996; 1999) studies, this study also defines readers’ beliefs as readers’ implicit models of reading, which consist of systematic beliefs about the readers’ perceived role as a reader. Schraw and Bruning (1996; 1999) proposed two different models, that is, transmission and transaction models. Each model consists of a variety of interrelated beliefs that constitute an internalized model of reading. Their framework was also adopted in this study, in which the transmission model is characterized by the belief that meaning is transmitted from the author and/or text; and the transaction model, by the beliefs that meaning is constructed by a transaction among the reader, author, and text. Reading Strategies Concerning the definitions for strategy, researchers in the field of language teaching and learning have many to offer. O’Malley and Chamot (1990) viewed -6-.

(17) learning strategies as the special thoughts or behaviors that individuals use to help comprehend, learn, or retain new information. Alexander, Graham, and Harris (1998) identified six attributes of strategies and stated that strategies are “procedural, purposeful, effortful, willful, essential, and facilitative in nature” (p. 130). Oxford (1990) defined learning strategies as specific steps taken by learners to enhance their own learning or make learning more effective. In addition, strategy is often distinguished from skill in that “a skill is an ability which has been automatized and operates largely subconsciously, whereas a strategy is a conscious procedure carried out in order to solve a problem” (William & Moran, 1989, cited in Urquahart & Weir, 1998, p. 98). Based on these definitions, the present study defined reading strategies as conscious acts learners take to help comprehend information or solve problems during the course of reading. Purpose of the Study The purpose of this study was to examine learners’ beliefs about reading and reading strategies in English as a foreign language and in their mother tongue, Mandarin Chinese. A group of Taiwanese college students with limited English proficiency formed the focus of the present inquiry. The primary concern was to investigate whether these learners’ beliefs about reading relate to their use of reading strategy in L1 (Chinese) and L2 (English) when they read for academic purposes. Readers’ beliefs were operationalized in terms of the two implicit models of reading, the transaction and transmission beliefs. Four research questions were addressed in the study: 1. What are learners’ beliefs about reading in L1 and L2 in terms of the transaction and transmission models of reading? Are there any differences in their belief systems between L1 and L2? 2. What are learners’ reported use of reading strategies in L1 and L2? Are there -7-.

(18) differences in their reported use of reading strategies between L1 and L2? 3. Are learners’ beliefs about reading in L1 and L2 related to their reported use of reading strategies? If so, do learners with distinctive beliefs about reading in both languages differ in their reported use of reading strategies? 4. Are learners’ beliefs about reading related to their actual use of reading strategies when they approach academic-oriented reading tasks in L1 and L2? In particular, do learners with distinctive beliefs about reading in both languages use different reading strategies? The first question aimed to compare differences in learners’ beliefs about reading between L1 and L2. The second inquiry intended to compare students’ reported use of reading strategies in L1 and L2. A quantitative approach with the use of questionnaires was adopted to address the above two questions. The data obtained from the questionnaires were also used to answer the third research question regarding the relationships between students’ beliefs about reading and their reported strategy use. More specifically, the study examined students with distinctive beliefs about reading in L1 and L2 to see whether differences existed in their reported pattern of strategy use. The final research question was proposed to further explore, using a qualitative approach, the relationships between readers’ beliefs and use of reading strategies. It aimed to see whether learners with distinctive beliefs about reading in the two languages show any differences in their actual processing of academic texts. Significance of the Study The significance of the study can be viewed from its contributions to the field of SLA in a broader sense as well as to the existing L2 reading theories and pedagogies in specific. In the field of SLA, although researchers have begun to recognize the importance of affective conditions in language learning, the affective domain of L2 learning still receive less attention compared with cognitive and -8-.

(19) linguistic areas (Yamashita, 2007). The current study, focusing on the relationships between learners’ beliefs about reading and their use of reading strategies, can provide additional evidence for research on affective factors in SLA. In the field of L2 reading research, the study of affect remains relatively unexplored. The study of beliefs about L2 reading in particular has been largely neglected given the small number of studies conducted in this area (Devine, 1988; Kamhi-Stein, 2003; Yang, G., 2004). The present inquiry hence fills this gap by investigating the possible impact of readers’ beliefs on their use of strategies. In addition, the study also offers a new direction for the research into L2 readers’ beliefs through the incorporation of Schraw and Bruning’s theory of transmission and transaction beliefs into a cross-linguistic investigation of readers’ beliefs. The findings of this study may advance our knowledge about the affective as well as processing differences between L1 and L2 reading, the impact of affective factors on reading processes, and the nature of reading across languages. Furthermore, the findings can also shed some light on the L2 reading instruction, particularly in EFL contexts similar to the one in Taiwan.. -9-.

(20) CHAPTER TWO. LITERATURE REVIEW. This chapter presents an overview of literature relevant to this study. The first section reviews important L1 reading theories in order to provide a theoretical background for the investigation of readers’ beliefs. The second section describes two lines of belief research conducted in the L1 research domain, which directly informed the type of reading beliefs examined in this study. The third section is concerned with the L1 reading strategy research. Finally, since this study targeted on a group of EFL/L2 readers, the last part of this chapter discusses three areas of L2 studies, learner beliefs in SLA, reading beliefs in L2, and reading strategies in L2. Major L1 Reading Theories In this section, I will discuss several major L1 reading theories, which can be divided into three parts. In the first part, three often-cited cognitive models of reading representing a more traditional and strictly cognitive view toward the reading process will be briefly reviewed. After that, I will begin to introduce one influential model of reading, the transactional theory of reading, put forth by Rosenblatt (1969, 1994). The theories discussed in the first two parts were intended to provide a theoretical basis for the two types of reading beliefs (i.e. the transmission versus transaction beliefs) being investigated. Finally, I will review two affective models of reading proposed by Mathewson (1994) and McKenna (1994) to explicate the importance of beliefs during the process of reading. Three Often-Cited Cognitive Models of Reading The Bottom-Up Approach The basic tenets of the so-called bottom-up approach of reading is that reading in general follows a linear pattern in which “readers create a piece-by-piece mental - 10 -.

(21) translation of the information in the text” (Grabe & Stoller, 2002, p.32). In other words, the analyses begin with the text, or bits of the text (Urquhart & Weir, 1998). The most frequently cited model is of Gough (1972). It represents a bottom-up view of reading in which reading process proceeds in a strict order from the letters, words, and larger linguistic units, to the interpretation of meaning. Another bottom-up model that has attracted much attention as well as criticism is the LaBerge-Samuels (1972) model of automatic information processing in reading. The notion of “attention” is central to this model, which has three characteristics: alertness, selectivity, and limited capacity. The reading process consists of two stages: first the decoding of the printed words; second, the comprehension of the decoded words. The theory helps to explain the differences between fluent and beginning readers in that in fluent reading, since the decoding process is done automatically, attention is then available for getting meaning from the words. According to the theory, through practice, a person develops skills at reading and then the processing of information can be performed with automaticity. Despite of the recent modification (Samuels, 1994), the model itself was still considered heavily text-based in nature because of its lack of concerns about meaning (Bernhardt, 1991). Although the bottom-up approach has suffered from its limitation to linear pattern of processing as well as lack of attention to meaning, in recent years, research has found that word-level processing, or decoding, plays a crucial role in reading comprehension (Pressley, 2000). Evidence accumulated from the vocabulary studies also supports that more extensive vocabulary promotes comprehension and that word identification skills are extremely important for fluent reading (Beck, Perfetti, & McKeown, 1982; Pressley, 2000). The Top-Down Approach Unlike the bottom-up approach which begins with the smallest unit of a text, - 11 -.

(22) the top-down approach begins with the reader. In other words, it characterizes the reader as someone who has a set of expectations about the text and samples information from the text to confirm or reject the hypotheses (Grabe & Stoller, 2002). The Psycholinguistic Guessing Game Model of Reading proposed by Goodman (1986) is an example of such approach. This model portrays reading as a process of “hypothesizing, sampling, and confirming information based on background knowledge, expectations about the text, a sampling of surface structures of the text and context information from the text ” (Grabe & Stoller, 2002, p.34). Although the model has been well-accepted among applied linguists, in recent years, it has received much criticism. One of the most damaging attacks is made about the claim that good readers guess more and use the context more than poor readers (Urquhart & Weir, 1998). Research has shown that good readers normally do not guess what words will appear next in the text and they tend to make less use of context in fluent reading (Grabe & Stroller, 2002). Moreover, it is generally easier for a skilled reader to simply recognize a text than try to generate predictions (Samuels & Kamil, 1988). In addition, research into eye movements also shows that in fluent reading, native speakers of English do look at most of the words on a page (Bernhardt, 1991) rather than only sampling parts of a text. The Interactive Approach A third approach, often termed the “interactive” model of reading, was first put forth by Rumelhart (1977, 1994). The essence of the theory is that it allows processing at one level or stage (e.g. word perception) to interact with processing at another level or stage (e.g. syntactic knowledge) (Carrell, Devine, & Eskey, 1988). Rumelhart argues that a skilled reader should be able to make use of sensory, syntactic, semantic, and pragmatic information to accomplish a reading task. Although the model also has some weaknesses, it is considered superior to the strictly bottom-up or - 12 -.

(23) top-down models because it “takes into account the critical contributions of both lower-level processing skills and higher-level comprehension and reasoning skills” (Grabe, 1991, p. 386). Another model that is often categorized within the interactive framework is the Interactive Compensatory Model proposed by Stanovich (1980). The key concept of this model is that a process at any level can compensate for deficiencies at any other level. More specifically, the model argues that “reading difficulties lead to increased interaction and compensation even among processes that would otherwise be more automatic” (Grabe & Stoller, 2002, p. 35). For instance, using context to understand a text or a word better is a compensatory strategy when encountering difficult words or texts. Rosenblatt’s Transactional Model of Reading Unlike the previously cited models, the transactional theory of reading (Rosenblatt, 1969; 1994) treats reading as a transaction and a composing activity. It is a two-way process, which involves three aspects: the reader, the text, and the meaning, each being an aspect of a relationship occurring at a particular time under particular circumstances. In this theory, the reader is an active participant in constructing meanings out of the text and brings in his or her own “linguistic-experiential reservoir” to the interpretation of the text. Meaning does not solely lie in the text but comes into existence during the transaction between the reader and the text. Rosenblatt also proposes the Efferent-Aesthetic continuum in which readers’ stance is positioned. Adopting an efferent stance means paying more attention to the public meaning or sense of a text which refers to “the cognitive, the referential, the factual, the analytic, the logical, the quantitative aspect of meaning” (Rosenblatt, 1994, p. 1068). The aesthetic stance, on the other hand, pays more attention to the more private, “sensuous, the affective, the emotive, the qualitative part of meaning” (Rosenblatt, 1994, p. 1068). - 13 -.

(24) The transactional model of reading has greatly influenced Schraw and Bruning’s (1996; 1999) conceptualization of readers’ implicit models of reading, transaction and transmission beliefs, which is also the focus of the present research. The reading models cited in the previous sections have largely described the process of reading in terms of the transmission dimension. In other words, the focus was on describing how readers decoded meanings from the text, which provided sources for meaning extraction. Rosenblatt’s model, however, emphasized a more active role of the reader during the process of meaning-construction. Following his view, knowledge, or meanings, did not come from the text alone but come into being when readers actively transact with the text. The theory provides theoretical basis for the distinction of two belief systems investigated in present study. In the next part, I will begin to briefly discuss two reading models which emphasize the importance of affective factors including the notion of beliefs during the reading process. Two Affective Models of Reading In reading research, concerns about the role of affective factors during the course of reading are less apparent. This is largely because most of the reading models were built by cognitive psychologists interested in the reading process (McKenna, 1994). It is not surprising that the three often-cited cognitive models, as described earlier, have not considered affective factors operating during the course of reading. A few theorists, however, have attempted to incorporate affective variables into their models. Mathewson (1994), based on the extension of his earlier model (1985), proposed the Model of Attitude Influence Upon Reading and Learning to Read, in which affective factors interact with cognitive processes during the act of reading. In this model, attitude is one of three principal factors influencing one’s intention to read; the others are external motivators and one’s emotional state. The intention to read then leads to reading behaviors in terms of text selection, attention, strategy use, and - 14 -.

(25) comprehension.. Another. important. aspect. of. Mathewson’s. model. is. its. three-component definition of attitude, which includes prevailing feelings about reading, action readiness for reading, and evaluative beliefs about reading. By treating attitude, an affective variable, as a major influence during the reading process, Mathewson made a considerable contribution to the study of affect in reading. Mathewson’s ideas were taken up by a number of writers and researchers, most notably McKenna (1994), who proposed a model of reading attitude acquisition. One of the major assertions McKenna made in his model was that attitude toward reading is directly influenced by two kinds of beliefs—beliefs about subjective norms as well as beliefs about the outcomes of reading. Therefore, in this model, beliefs about reading are causally related to the development of attitude toward reading. Furthermore, the model also specifies the possibility that attitude may be altered by the results of a specific reading experience. The change of attitude can be influenced through a direct route involving the immediate impact from the specific act of reading and an indirect route from the experience of reading to the alterations in beliefs about the outcomes of reading. It should be noted that although Mathewson and McKenna’s models primarily focused on the importance of reading attitude during the process of reading, both models were found to incorporate beliefs about reading as an important component on the formation of reading attitude. By including the various notions of beliefs in the construct of reading attitude, these theorists seemed to recognize the essential role of beliefs during the act of reading. It is therefore particularly meaningful for the present study to explore the notion of beliefs in the cross-linguistic reading processes. In the above section, I have described the three strictly cognitive models of reading and also the transactional model as a theoretical basis for the contrast between the transmission and transaction beliefs about reading. I have also illustrated the - 15 -.

(26) importance of the notion of beliefs during the process of reading by drawing on the two affective reading models found in the literature. In the following part, I will begin to review studies on learner beliefs conducted in the first language domain. Reading Beliefs in L1 In this section, two lines of research into reading beliefs will be discussed: Schommer’s (1990, 1994) work of epistemological beliefs and Schraw and Bruning’s (1996, 1999) study of readers’ implicit models of reading. Schommer (1990) investigated students’ beliefs about the nature of knowledge and learning, or epistemological beliefs, in an attempt to understand the contributing factors to poor learners’ learning problems. Schommer (1990) developed a questionnaire to test the multidimensionality of an epistemological belief system. Statistical analysis of the study yielded four dimensions of the beliefs about the nature of knowledge and learning: (1) fixed ability, (2) simple knowledge, (3) certain knowledge, and (4) quick learning. In addition, the results from the various measures of text comprehension showed that the less college students believe in simple knowledge, the better they perform on the mastery test and the more accurately they assess their own understanding. Furthermore, it is also found that the less students believe in simple knowledge, the more sophisticated study strategies they reported using. According to Schommer (1994), research evidence also suggested that strong beliefs in quick learning, simple knowledge, and fixed ability hinder learning as well. Schommer (1994) further modified the earlier theory and proposed that an epistemological belief system is multidimensional, with the four beliefs dimensions more or less independent from each other. The four beliefs include the Source, Structure, Certainty of Knowledge, and the Speed and Control of the Acquisition of Knowledge. In her summary of research conducted on the topic of epistemological beliefs, Schommer (1994, p.38) concluded that “there is evidence that suggests - 16 -.

(27) epistemological beliefs affect comprehension, metacomprehension, interpretation, and persistence.” With their specific focus on beliefs about reading, Schraw and Bruning (1996) conducted a study to examine the relationship between readers’ implicit models of the reading process and reader engagement. They developed an instrument, called the Reader Belief Inventory, which was tested on a group of university students in the U.S. Based on the results of factor analyses, Schraw and Brunning identified two models of beliefs, the transmission and transaction models. The transmission model is characterized by the beliefs that meaning is transmitted from the author and/or text and is largely independent of the reader, whereas the transaction model is based on the assumption that meaning is constructed by a transaction between the reader, author, and text. Using the Reader Belief Inventory, Schraw and Brunning were able to classify the students into different reader groups. Participants of their study were asked to read an 800-word text and complete two comprehensions measure tests including a free-recall test and a reader response essay. The results of their study revealed two things: First, those who endorsed a transactional model had higher recall of the expository essay; second, those who were classified as high on the transaction dimension reported more critical evaluations of the text, were more likely to relate text information to prior knowledge, and showed more affective responses. In a follow-up study, Schraw and Bruning (1999) further investigated the relationship between readers’ implicit models of reading and their motivation to read. They argued that adult readers possess multiple belief systems that shape the ways they read. A strong commitment to the transaction model increases motivation to read by strengthening the degree to which readers value the meaning-construction process and by increasing the type and number of deeper processing strategies. In short, both Schommer’s work of epistemological beliefs and Schraw and - 17 -.

(28) Bruning’s study of readers’ implicit models of reading have provided important bases for the current investigation on EFL readers’ beliefs and reading strategies. Their studies have shown that learner beliefs about the nature of knowledge and learning have a considerable impact on their comprehension of a text. Additionally, readers’ implicit models of the reading process or assumptions about their role of the reader can have an effect on reading comprehension, reading engagement, and motivation to read. However, since these studies were conducted in the first language context, it is not clear whether the arguments proposed by these researchers could be extended to a second or foreign language setting. Moreover, as described in the first chapter, one important limitation of Schraw and Brunning’s studies (1996; 1999) is that they only examined the effect of beliefs on readers’ comprehension products (i.e. the response essays). Such investigation by nature was insufficient in providing evidence for whether and how the endorsement of certain beliefs might have an influence on one’s reading processes. It is therefore particularly significant for the present study to focus the investigation on the learners’ use of reading strategies in order to have a more comprehensive understanding towards this issue. Reading Strategies in L1 Since the study of L2 reading strategies has been greatly influenced by strategy research conducted in the first language domain, the following review will describe two lines of L1 strategy research that has also been well-received in the field of L2 reading. One major focus of the early comprehension research in L1 was the investigation of strategies used by good readers. Many of the published studies on good readers’ comprehension processes conducted in the 1970s and 1980s were reviewed in Pressley and Afflerbach’s (1995) book, in which 38 think-aloud studies were analyzed. They grouped reading strategies (or activities) into three types: identifying strategies, which help in constructing the meaning of the text; monitoring - 18 -.

(29) strategies, which serve to regulate comprehension and learning; and evaluating strategies, by which readers reflect or respond in some way to the text. Based on their analysis, Pressley and Afflerbach were able to characterize the reading processes of good readers. Duke and Pearson (2002, p. 205-206) summarized Pressley and Afflerbach’s (1995) and Block and Pressley’s (2001) research work on good readers’ comprehension and provided a list of what good readers do when they read: (1) Good readers are active readers. (2) They have clear goals in mind for their reading. (3) Good readers typically look over the text before they read. (4) As they read, good readers frequently make predictions about the text. (5) They read selectively. (6) Good readers construct, revise, and question the meanings they make as they read. (7) Good readers try to determine the meaning of unfamiliar words and concepts in the text. (8) They draw from, compare, and integrate their prior knowledge with material in the text. (9) They think about the authors of the text. (10) They monitor their understanding of the text. (11) They evaluate the text’s quality and value and react to the text in a range of ways, both intellectually and emotionally. (12) Good readers read different kinds of text differently. (13) When reading narrative, good readers attend closely to the setting and characters. (14) When reading expository text, good readers frequently construct and revise summaries of what they have read. (15) For good readers, text processing occurs not only during “reading” but even after the “reading” has ceased. (16) For good readers, comprehension is a satisfying and productive activity. In addition to characterizing good readers’ use of reading strategies, another line of L1 comprehension research focused on comparing the strategies used by good and poor readers (Kletzien, 1991; Olshavsky, 1976-1977). Similar to the studies of good readers, researchers have typically collected self-report data from protocols or interviews to examine strategy use by readers with different reading abilities. - 19 -.

(30) Kletzien’s (1991) study represents this research trend. In her study, Kletzien examined how the interaction of reader ability, text difficulty, and strategy knowledge affects regulation of strategy use among good and poor comprehenders. A total of 48 high school students, who were half good comprehenders and half poor comprehenders, read three expository passages of increasing difficulty. Students were asked to complete a cloze task for each passage and to explain their reasoning processes. Students’ self-reports of their thinking processes were collected and analyzed to identify their comprehension strategies. A total of 15 categories of comprehension strategies were identified and included in Kletzien’s classification scheme. These included: using syntax or punctuation, using the author’s style, using known phrases, rereading previous text, reading subsequent text, recognizing the structure of passage or paragraph, recognizing the structure of the sentence, using prior knowledge, visualizing, using the main idea, making an inference or drawing conclusions, paraphrasing, looking for key vocabulary or phrases, no awareness of strategies, and non-use of strategies. The results of the study showed that all subjects reported depending heavily on using key vocabulary, rereading, making inferences, and using previous experience in constructing responses for all three passages. As for the differences between groups (i.e. the good versus poor readers), it was found that the two groups used the same type and number of strategies on the easy passage, but as the passage difficulty increased, good comprehenders used more types of strategies and used strategies more often than the poor comprehenders did. In sum, grounded in the studies of good readers, L1 comprehension research has provided a rich source for the study of reading strategies in L2. Moreover, the research into the comparison of strategy use by good and poor L1 readers has also shed some light on the studies of L2 readers with different language and reading abilities. The last section of this chapter will provide a review of research on L2 - 20 -.

(31) reading strategies. Learner Beliefs in SLA Research into learner beliefs in the field of SLA emerged from the late 1980s. It began with the two pioneer works of Wenden (1987) and Horwitz (1987). Wenden (1987), in her interviews with a group of 25 adult ESL learners, investigated student beliefs or views about the nature of learning a second language. Twelve explicit statements, representing learners’ prescriptive beliefs, were identified in her study and categorized into three main groups: beliefs about the importance of using the language; beliefs about the need to learn about the language, especially grammar and vocabulary; and beliefs about the role of personal factors. In addition, Wenden, also based on the interview protocols, provided preliminary evidence that student beliefs about language learning can influence their approach to language learning in terms of strategies, attention pattern, evaluation criteria, and planning priorities. In a similar vein, Horwitz (1987; 1988) explored learner beliefs first through eliciting responses from a group of English and foreign language teachers, educators, as well as students. Based on their responses, she then took a step further to develop an instrument, known as the Beliefs About Language Learning Inventory (BALLI), to assess student opinions on issues related to language learning. The BALLI contains thirty-four items and assesses student beliefs in five major areas: difficulty on language learning; foreign language aptitude; the nature of language learning; learning and communication strategies; and motivations and expectations. Horwitz (1987; 1988) reported that individuals often hold different beliefs or notions about language learning and these preconceived beliefs have the potential to influence both their learning experiences and actions as language learners (Horwitz, 1999). Influenced by these two pioneer works, particularly with the BALLI questionnaire developed by Horwitz, a considerable number of studies have been conducted in the - 21 -.

(32) field of SLA concerning student beliefs about language learning and their connections to various aspects of learning. In addition to Horwitz’s BALLI questionnaire, other SLA researchers also made similar attempts to develop instruments that helped identify and classify learner beliefs. Cotterall’s (1999) and Sakui and Gaies’s (1999) studies represented such efforts. Similar to BALLI, these instruments of beliefs all incorporated a variety of issues related to language learning. One of the issues being investigated was the use of learning strategies. In following part, I will review studies that examined the relationship between learner beliefs and various aspects of language learning, including the use of learning strategies. Yang (1999) conducted a study to examine the relationship between EFL learner beliefs and the use of language learning strategies. In her study, information on learner beliefs and use of language strategies was collected from 505 Taiwanese university students through an instrument called English Learning Questionnaire, which was composed of Horwitz’s (1987) BALLI, Oxford’s (1990) Strategy Inventory for Language Learning (SILL), and author-designed open-ended questions. The study showed that students’ self-efficacy beliefs about learning English were strongly related to their use of all types of strategies, especially the functional practice type. Moreover, students’ beliefs about the value and nature of learning spoken English were closely linked to their use of formal oral-practice strategies. Based on the results, Yang further proposed a theoretical construct of language learning beliefs to capture the cyclical relationships among learner beliefs, motivation, and the use of strategies. In addition to strategies, leaner beliefs were also found to be related to the outcomes of foreign language learning. Mantle-Bromley (1995) conducted a study to investigate the relationship among learner attitudes, beliefs, and learning outcomes. Participants were middle-school students in the U.S. participating in a 9-week foreign - 22 -.

(33) language program which was designed with a particular attempt to improve students’ attitude toward the native speakers of French and Spanish. The results of her study demonstrated that young students with misconceptions about language learning from the outset might hinder their progress and persistence in language study. With a similar focus on leaning outcomes, Huang (2003) explored the relationship between learner beliefs and proficiency levels. The study investigated the differences of the beliefs high English-proficiency and low English-proficiency level learners hold in English learning. Data were collected from a group of Taiwanese high school students through the use of two instruments (the BALLI questionnaire and GEPT, a locally developed standardized English proficiency test) and student interviews. The results of her study revealed five significant belief discrepancies between the high- and low-proficiency groups. These include beliefs in language aptitude, the difficulty of language learning, self-efficacy, nature of language learning (the importance of translation), and the use of communication strategies. In short, several points can be made from the above review of studies. First, it seems that learners often hold certain beliefs about language learning. Second, these preconceived beliefs often tie closely to learners’ choices of language strategies. Third, unrealistic beliefs or misconceptions about language learning may have a negative impact on students’ progress in language study. Furthermore, there appears to be some discrepancies in beliefs between high and low proficiency learners groups. Reading Beliefs in L2 Although a considerable effort has been made to the study of learner beliefs about language learning in general, only a few studies addressed the issue of students’ beliefs about reading in a foreign/second language. In this section, I will review several studies conducted with a specific focus on beliefs or attitudes about L2 reading. - 23 -.

(34) Devine (1988) conducted a case study of two ESL readers to investigate the internalized models of reading and the relations to reading behaviors. Her study drew on the classic work of Haste and Burke’s (1977) distinction of three models of reading processes: the sound-, word-, and meaning-centered models. Her data were collected from the subjects’ oral reading interviews, oral reading samples, and retellings of the oral reading. The results showed that the two learners, one identified as sound-centered the other as meaning-centered, had quite distinct reading behaviors. The sound-centered learner in her study appears to over-rely on the grapho-phonic cueing system to the extent of sacrificing the comprehension of meanings from the text. Devine (1988) concluded that ESL readers’ theoretical orientations toward reading may influence the ways in which they approach a text and their comprehension of that text. A study done by Auerbach and Paxton (1997) is also related to learners’ views of reading though the major concern of it was to investigate the effect of having students reflect and articulate their individual use of strategies on their reading comprehension. At the initial stage, the study used the conception questionnaire adapted from Devine (1988) to get students’ views about reading in English. According to the author, the initial data revealed that students who attributed their difficulty in L2 reading to their lack of L2 proficiency seemed to have a sense of insecurity about their reading and have negative feelings about reading in English. Also focusing a small group of ESL readers, Kamhi-Stein’s (2003) conducted a study to explore the relationship between L1 and L2 reading strategy use and affective factors. His investigation focused on two affective factors, i.e., readers’ views of their native language and their beliefs about reading. The participants consisted of four under-prepared L2 college readers of Spanish and English. Data were collected through think-aloud protocols, interviews, and reading comprehension measures. The - 24 -.

(35) results of the study were found to largely confirm Devine’s observations in which readers who viewed reading as a process of meaning construction tended to be more multi-strategic, while those who viewed reading as primarily word-centered tended to be logo-centric during reading. A more recent attempt to the study of learner beliefs about reading comes from the work of G. Yang (2004). The study used a qualitative method to explore the relationship between readers’ beliefs and reading strategies. The subjects were seven Chinese students studying in the United States. Data were collected through interviews and think-aloud protocols. Findings of the study suggested that students do hold certain beliefs about reading in English. Some of these beliefs included: (1) It was important to read in English for comprehension for academic purposes; (2) Reading extensively in English would help expand a reader’s English vocabulary and improve a person’s overall English language skills; (3) Reading in English was also for information and for pleasure; (4) Reading comprehension was based on the reader’s linguistic knowledge as well as other knowledge sources such as the reader’s background knowledge; and (5) Good readers used different strategies to read with good comprehension and at a fast rate. In addition, the study also showed a close relationship between the participants’ beliefs about reading in English and their use of reading strategies. In addition to the study of beliefs, the study of other affective factors such as attitudes has also been found in the field of L2 reading. Yamashita (2007) examined the relationship of reading attitudes between L1 and L2. She focused on Japanese EFL learners’ reading attitudes. Based on her earlier work (2004), a reading attitude questionnaire was developed to investigate the issue. Using factor analysis, five major components were found in the scale including comfort, intellectual, practical value, anxiety, and linguistic. In the study, she found that first adult EFL learners have - 25 -.

(36) developed distinctive attitudes toward reading in L1 and L2. These differences included that students were found to feel more comfortable in reading L1 than L2; students believe that they achieve higher intellectual development from reading in L1 than L2; students believe that they obtain higher practical benefits from reading L2 than L1; students feel more anxious in reading L2 than L1; and students believe they achieve higher linguistic development from reading L2 than L1. Second, through the multiple regression analyses, she found that there was a transfer of reading attitudes from L1 to L2. In addition, the results of her study also demonstrated that L2 proficiency seemed to play a much weaker role in the transfer of reading attitudes than that of reading ability. In summary, research into L2 readers’ beliefs suggests that learners appear to have certain theoretical orientations toward L2 reading or hold certain beliefs about L2 reading. Moreover, the research evidence also indicates a close relationship between readers’ beliefs and their use of reading strategies. Finally, the studies of reading attitudes reveal that although students tend to have different attitudes toward reading in L1 and L2, they also tend to transfer their reading attitudes from L1 to L2. Reading Strategies in L2 The following two sections provide a review of previous L2 research devoted to the investigation of reading strategies. Based on the type of research method being used, these studies of L2 reading strategies were sorted into two groups: the think-aloud studies and the questionnaire studies. Studies Using Think-Aloud The Think-Aloud Method Since the think-aloud method has been widely used to study the reading process in both first and second language domains, it is necessary to understand its basic tenets and potential benefits. Generally speaking, this method requires the readers to - 26 -.