作者：苗丽霞
出版社：上海交通大学出版社
格式: AZW3, DOCX, EPUB, MOBI, PDF, TXT
词汇知识、工作记忆与二语阅读理解的关联性研究(英文版)试读：
前言
近期的阅读理解研究多基于阅读理解多成分观理论视域。根据该理论，阅读理解是一种包含多种认知和语言技能成分的复杂认知过程。阅读理解过程的解读不仅需要探究各个阅读能力成分的单独作用，还需要了解其互动机制。本书聚焦于词汇知识和工作记忆这两个重要的阅读能力成分，考察它们对中国成人英语学习者阅读理解能力的单独作用及其交互性影响。
本书的论述主要基于以下三个方面的原因。第一，国内外同类研究多聚焦于词汇知识或者工作记忆对阅读理解的单独贡献，同时包含两个变量且对其单独和交互性作用进行综合性考察的多维度研究尚不多见。第二，虽然国内外有关词汇知识和阅读理解关系的研究较为丰富，然而由于词汇知识的多维性特征，其与阅读理解的关联本质仍需进一步探究。一方面，关于词汇知识深度在阅读理解中的作用，已有的研究结论不尽一致；另一方面，已有研究多聚焦于词汇知识的陈述性范畴，即词汇知识宽度和词汇知识深度，而对于词汇知识的程序性范畴，即词汇知识流利度（词汇知识的提取速度）鲜有关注。第三，二语阅读研究中，关于工作记忆和阅读理解关系的研究数量不多，且存在一定的局限性。多数研究要么将阅读理解表征为一个整体构念，忽略其层级性本质特征，要么仅考察工作记忆容量对阅读理解的影响，没有深入探究工作记忆的加工和存储两个成分对阅读理解的具体作用。
本书以中国外语语境下中级水平英语学习者为受试对象，采用纸笔测试和反应时实验等研究方法，运用相关和回归等统计手段，揭示二语词汇知识、工作记忆和二语阅读理解三个变量间的关联机制。研究考察的三个词汇知识维度分别是：词汇宽度（学习者所认识词汇的数量）、词汇深度（学习者所掌握词汇知识的质量）和词汇流利度（学习者识别单词的速度）。其中，词汇深度涵盖两个方面：词汇语义深度和词法深度。前者指学习者对单词的横组合和纵聚合语义关系的掌握程度；后者指学习者对词汇结构知识的了解程度。阅读理解的两个层次分别是：字面阅读理解和推断性阅读理解。本书主要有如下三个方面的研究结果：第一，虽然词汇知识的三个维度均与阅读理解成绩显著相关，但仅有词汇宽度和词汇语义深度对阅读理解能力有直接贡献，词法深度和词汇流利度通过词汇宽度或者词汇语义深度间接地作用于阅读理解能力，相比较而言，三个词汇知识维度中，词汇宽度知识对阅读理解能力的预测力最强；第二，工作记忆与阅读理解能力和推断性阅读理解能力显著相关，但与字面阅读理解能力不具有相关关系；在工作记忆的两个构成成分中，加工成分与阅读理解能力和推断性阅读理解能力显著相关，但与字面阅读理解能力不相关；第三，词汇宽度和词法深度与工作记忆在阅读理解过程中具有负向的交互作用，学习者在阅读理解过程中对工作记忆的依赖程度依其词汇水平而异，词汇水平低的学习者对工作记忆的依赖强度更大。
本书的研究结果兼具理论和实践意义。理论方面，揭示了二语阅读理解中词汇知识的多维性本质特征，阐明了工作记忆及其加工和存储成分在字面和推断性阅读理解过程中的作用，揭示了二语阅读理解过程中语言和认知成分间的互动机制。实践方面，本书为二语阅读教学提供了教学启示，阅读教师需要关注阅读理解过程中多维度词汇知识的不同影响以及学习者个体因素在其间的协变作用。
本书的写作得到多位老师、同事和家人的帮助与支持。感谢马广惠教授在确定研究选题和实施过程中的指导和鼓励，以及在数据统计分析方面的专业指引！感谢丁言仁教授、倪传斌教授、濮建忠教授、汪少华教授和张辉教授在选题论证过程中的宝贵建议！感谢我的同事励哲蔚老师、丛迎旭老师、戚国辉老师、盛盈老师、林渭芳老师和徐燕老师在数据收集过程中的无私奉献！感谢我的家人对我的理解和支持！List of Abbreviations
AWL Academic Word List
BNC British National Corpus
CC READER Capacity Constrained Reader
CET-6 College English Test Band 6
DM declarative memory system
DP declarative/procedural
DVK depth-of-vocabulary-knowledge
EFL English as a foreign language
ELLs English language learners
EVST European Vocabulary Size Test
ESL English as a second language
GSL General Service List
L1 first language
L2 second language
LISREL Linear Structural Relations
MA master of arts
MK morphological knowledge
MS master of science
NETEM National Entrance Test of English for MA/MS Candidates
PM procedural memory system
RC reading comprehension
RCL literal reading comprehension
RCI inferential reading comprehension
SAT Scholastic Aptitude Test
SEM Structure Equation Modeling
TEM 4 Test for English Majors Band 4
TOEFL Test of English as a Foreign Language
TOEFL iBT TOEFL Internet Based Test
VB vocabulary breadth
VBhih high-level vocabulary breadth
VBlw low-level vocabulary breadth
VD vocabulary depth
VF vocabulary fluency
VKS Vocabulary Knowledge Scale
VLT Vocabulary Levels Test
VMD vocabulary morphological depth
VMDhih high-level vocabulary morphological depth
VMDlw low-level vocabulary morphological depth
VS vocabulary size
VSD vocabulary semantic depth
VST Vocabulary Size Test
WAF Word Associates Format
WM working memory
WMP the processing component of WM
WMS the storage component of WM
WPLT Word Part Levels TestList of Tables
Table 1 The demographic information of the participants（N=67）/76
Table 2 An example of the test format of the VST/78
Table 3 An example of the format of the WAF/79
Table 4 An example of the meaning section of the WPLT/82
Table 5 Variables for the statistical analyses in the study/90
Table 6 Means, standard deviations and score ranges on RC, VB, VSD, VMD and VF/98
Table 7 Pearson correlations among RC, VB, VSD, VMD and VF/99
Table 8 Regression results with VB, VSD, VMD and VF as the independent variables/103
Table 9 Regression results with VB and VSD as the independent variables/104
Table 10 Indirect effects of VMD and VF on RC/108
Table 11 Means, standard deviations and score ranges of RC, RCL, RCI, WMP, WMS and WM/115
Table 12 Pearson correlations among RC, RCL, RCI, WMP, WMS and WM/116
Table 13 Results of paired samples T-tests between RCL and RCI/117
Table 14 Pearson correlations among RC, WM, VB, VSD, VMD and VF/128
Table 15 Vocabulary mediating effects on the relationship between WM and RC/132
Table 16 Pearson correlation coefficients between WM and RC for the VB groups/134
Table 17 Pearson correlation coefficients between WM and RC for the VMD groups/134List of Figures
Figure 1 The multi-component model of working memory（Baddely, 2003）/22
Figure 2 The embedded-process model of working memory（Cowan, 2005）/23
Figure 3 The three dimensions of vocabulary knowledge/71
Figure 4 The diagram of mediation analysis/92Chapter OneIntroduction
The first chapter is to describe the background and the significance of the present study. The organization of the book will also be presented.1.1 Background of the Study
Recent investigations of reading comprehension in both first (L1) and second (L2) language contexts have often adopted the component-skills approach, which views reading comprehension as a multivariate skill involving a complex combination and integration of a variety of cognitive, linguistic, and meta-linguistic processes. A better understanding of reading comprehension relies not only on the investigation of how much respective and collective contribution each of these processes makes to reading comprehension, but also on how different processes interact with each other.
Both vocabulary knowledge and working memory (WM) have been found to be key correlates of second language reading comprehension (Grabe, 2009；Jeon & Yamashita, 2014). Many empirical studies have demonstrated that there is a close link between vocabulary knowledge and second language reading comprehension (Laufer, 1989, 1992; Hirsh & Nation, 1992; Hu & Nation, 2000; Qian, 1999, 2002). Within this area of interest, however, most studies have often focused on the two dimensions of breadth and depth of vocabulary knowledge in second language reading comprehension (Laufer, 1989, 1992; Qian, 1999, 2002), and have not fully addressed the multidimensional nature of vocabulary knowledge (Kieffer & Lesaux, 2008). Vocabulary knowledge is a complex and multidimensional construct (Nation, 2001; Perfetti, 2007). In addition to a large vocabulary size, rich knowledge about each lexical item, and a well-organized mental lexicon where lexical items have numerous formal and semantic connections with each other, a person's vocabulary mastery also lies in the automaticity in which lexical items can be recognized and produced, which is a key requirement of real-time vocabulary use (Schmitt, 2010). However, there have been few studies investigating the effect of lexical automaticity on reading comprehension, especially in the area of L2 research. Even among the existing studies investigating the relative effects of breadth and depth of vocabulary knowledge on second language reading comprehension, serious inconsistencies exist in terms of the role of different components of lexical depth in reading comprehension. For instance, there has been much discrepancy concerning the contribution of morphological depth to second language reading comprehension among different researchers. Some researchers (e.g. Kieffer & Lesaux, 2008, 2012; Jeon, 2011) provided evidence that morphological awareness is a significant predictor of second language reading comprehension while others (e.g. Qian, 1999; Zhang & Koda, 2012) found that the contribution of morphological awareness to second language reading comprehension is negligible or indirect. It is, therefore, necessary to investigate how lexical fluency, the ability to retrieve lexical knowledge with ease and speed, works and interacts with other aspects of lexical knowledge in second language reading comprehension and also to resolve the controversies pertaining to the role of vocabulary depth in second language reading comprehension.
With its restricted functions of processing and storage, WM has also been found to play an important role in reading comprehension in numerous first language studies (see Daneman & Merikle, 1996, for a review). However, research on the relationship between WM and L2 reading comprehension is still scanty (Alptekin & Erçetin, 2011). Even among the small number of relevant studies, most either treat second language reading comprehension as a global construct without taking into account of the multilevel representational architecture of reading comprehension (e.g. Harrington & Sawyer, 1992; Leeser, 2007), or only investigate the effect of WM as a whole on second language reading comprehension without delineating the separate roles of the processing and storage components of WM in second language reading comprehension (e.g. Alptekin & Erçetin, 2009; Walter, 2004). Multidimensional studies, therefore, are required to further explore the complex relationship between WM and second language reading comprehension.
In terms of the effects of lexical knowledge and WM on reading comprehension, most studies have often focused on their respective contribution to reading comprehension. Few studies to date, especially in the area of L2 research, have investigated their collective contribution to reading comprehension or the interaction in their role in second language reading comprehension. Although there is a large body of knowledge about the influence of lexical knowledge and WM in both L1 and L2 studies (Alptekin & Erçetin, 2009, 2010, 2011; Cain, Oakhill, & Bryant, 2004; Daneman & Capenter, 1980; Laufer, 1989, 1992; Qian, 1999, 2002), less is known about how these two important reading correlates work together in the service of something as complex as reading comprehension, especially in the area of L2 studies. Given that factors influencing reading ability become increasingly interrelated across development (Stanovich, 1980), it is not unreasonable to assume that a better understanding of L2 reading comprehension processes will depend not only on the relationship between the predictors and reading outcomes, but also on the relationships among the predictors themselves (e.g. Hannon, 2012).
The present study, therefore, aims to explore both the separate contribution of L2 vocabulary knowledge and WM to and their combined effects on L2 reading comprehension. It is hoped that addressing these under-researched questions can, on the one hand, lead to a better understanding of L2 reading comprehension processes, and on the other hand, facilitate L2 reading comprehension pedagogy as well as L2 vocabulary acquisition.1.2 Significance of the Study
The investigation of the relationship among L2 lexical knowledge, WM and L2 reading comprehension will contribute theoretically, methodologically and pedagogically to the field of L2 reading comprehension and L2 vocabulary acquisition.1.2.1 Theoretical significance
The theoretical significance of the present study is threefold:
1) Providing empirical evidence to the interaction between L2 lexical knowledge and WM in their contribution to L2 reading comprehension.
Reading comprehension is a multi-component process, composed of multiple linguistic, cognitive, and meta-linguistic processes such as decoding, vocabulary knowledge, syntactic processing, meta-cognition, and WM (Grabe, 2009; Koda, 1996, 2005; Bernhardt, 2011). Many of these processes enjoy substantial bodies of research that has been directed at understanding how each component process operates. Almost all of the existing research on text comprehension has focused on identifying and examining the various component processes in isolation (Kintsch & Rawson, 2005). Such investigations have gone a long way furthering our understanding of how reading comprehension works. The nature of reading comprehension processes, however, is still far from clear. Text comprehension is not simply the sum of these various processes, but arises from their coordinated operation as a system. It is, therefore, theoretically significant to investigate not only the separate contribution of each of these components, but also their combined effects. Just as Kintsch and Rawson (2005) pointed out, an important direction for future research will be to examine the interaction of the various components to understand how they work together as a system to give rise to comprehension.
Although small in number, several attempts have been made by researchers to examine the coordinated relationship between vocabulary knowledge and WM in reading comprehension in L1 context. The first attempt, to the writer's knowledge, was made by Dixon, LeFevre, and Twilley (1988), who in evaluating the relative roles of word knowledge and working memory, found much correlation between the two independent variables in their causal model analyses, suggesting the combined and interactive influence of word knowledge and WM on reading comprehension. Chrysochoou, Bablekou, and Tsigilis (2011) also found the mediating role of vocabulary knowledge in several associations between WM and reading comprehension measures in Greek children. Rose and Rouhani (2012), in a sample of adolescents with dyslexia, demonstrated a negative interaction between verbal WM and expressive vocabulary in oral reading fluency.
Despite a lack of empirical research in L2 studies, the presence of an interaction between WM and vocabulary knowledge is consistent with several theories of reading, including the automaticity theory (Laberge & Samuels, 1974), the Interactive-Compensation model (Stanovich, 1980; Bernhardt, 2011), and the Construction-Integration model (Kintsch, 1988, 1998). This study, therefore, will not only provide empirical evidence for the interaction between L2 vocabulary knowledge and WM in L2 reading comprehension, but also have direct impact on our understanding of the nature of L2 reading comprehension processes.
2) Exploring more thoroughly the nature of the link between L2 vocabulary knowledge and L2 reading comprehension.
Vocabulary knowledge is a multidimensional construct. Meara (1996a) proposed a three-dimensional framework of vocabulary knowledge, consisting of breadth, depth and fluency of lexical knowledge. The first two dimensions (i.e. breadth and depth) belong to the category of declarative knowledge while the third one procedural knowledge (Milton & Fitzpatrick, 2014). Improvement of lexical mastery consists of both declarative knowledge and the ability to use this knowledge with speed and ease (Schmitt, 2010). According to the Access Hypothesis (Mezynski, 1983), lexical fluency is the crucial variable in the relationship between vocabulary knowledge and reading comprehension for it contributes most directly to reading comprehension. In other words, reading comprehension depends more on how quickly and easily a reader can access the correct meanings of words rather than how many words he knows or how much he knows about the words.
However, in spite of its theoretical importance, the dimension of lexical fluency has not often been included in previous empirical studies. In L1 studies, although Mezynski's (1983) review study suggests a facilitative effect of lexical automaticity on children's reading comprehension, Tannenbaum, Torgesen, and Wagner (2006) were not able to find any unique contribution of lexical fluency to reading comprehension; depth and fluency were not completely distinguishable from one another in their confirmatory factor analyses. In L2 studies, lexical fluency is rarely included as a dimension of lexical knowledge. Therefore, the picture to date is still far from clear in terms of the influence of lexical fluency on and its possible interaction with other aspects of vocabulary knowledge in reading comprehension. The present study aims to investigate the effects of the three dimensions of lexical knowledge, namely vocabulary breadth (VB), vocabulary depth (VD) and vocabulary fluency (VF), thus providing empirical evidence, which will, on the one hand, help to clarify the complex relationship between vocabulary knowledge and L2 reading comprehension, and on the other hand, shed light on the direct impact of lexical fluency on L2 reading comprehension.
3) Providing empirical evidence for a better understanding of the functioning mechanism of WM in L2 reading comprehension.
Working memory, with its dual functions of processing and storage, is a limited-capacity system that is necessary for complex cognitive tasks. Ever since the publication of the seminal paper by Daneman and Carpenter (1980), an extensive body of research has been done investigating the relationship between WM capacity and reading comprehension in both L1 and L2. WM capacity has been found to be a fundamental cognitive element in reading comprehension and correlates significantly with reading comprehension in both L1 and L2. However, in terms of its functioning mechanism in reading comprehension, controversies still exist. Some researchers (e.g. Daneman & Carpenter, 1980; Just & Carpenter, 1992; Waters & Caplan, 1996) argued that the processing component of WM determined its correlation with reading comprehension. The correlation between WM and reading comprehension was due to the overlap of operations across the processing component and reading comprehension tasks. However, the overlapping view has recently been challenged by some researchers (e.g. Bayliss, Jarrold, Gunn, & Baddeley, 2003; Unsworth, Thomas, Heitz, Broadway, & Engle, 2009), whose studies showed that the storage component of WM also made unique contribution to reading comprehension beyond the processing component. Their results add to the complexity of the link between WM and reading comprehension. Whether it is the processing component, the storage component, or both, that determines the correlation between WM capacity and reading comprehension is still in the air. Therefore, more research needs to be conducted to clarify the functioning mechanism of WM in reading comprehension.
The relationship between WM and reading comprehension can also be complicated by the multilevel nature of reading comprehension. Reading comprehension can be seen as involving distinct levels of text comprehension. Gray in 1960 distinguished between reading the lines, reading between the lines, and reading beyond the lines, indicating three different levels of text understanding ranging from literal understanding, inferential comprehension and critical comprehension (Alderson, 2000). Kintsch (1998) also described two levels of text comprehension, consisting of the text-base and the situational model of the text. However, research examining the relationship between WM and reading comprehension has often treated reading comprehension as a global construct, disregarding the multi-level representational architecture. This is a problematic issue in that WM may be differentially involved in different levels of text comprehension as different levels of text representation may involve distinct linguistic and cognitive processes (e.g. Cutting & Scarborough, 2006; Keenan, Betjemann, & Olson, 2008). Therefore, given the multiple levels of reading comprehension, it would be theoretically useful to clarify the relationship between WM and reading comprehension for a thorough understanding of the nature of their association.1.2.2 Methodological significance
The existing studies investigating the relationship between L2 vocabulary breadth and L2 reading comprehension often employ the Vocabulary Levels Test (VLT). The VLT was originally created by Nation (1983) and later updated by Schmitt, Schmitt, and Clapham (2001), for the purpose of measuring second language learners' vocabulary knowledge at five frequency levels, the most frequent 2,000, 3,000, 5,000, 10,000 word families and the Academic Word List (AWL) (Coxhead, 2000). Although it has often been misinterpreted as a measure of vocabulary size, it was originally designed to measure the extent of second language learners' vocabulary knowledge at each level, thus indicating where vocabulary learning was needed. As a measure of second language learners' vocabulary size, it has several limitations. Firstly, it is not a valid measure of vocabulary breadth because there are gaps in vocabulary knowledge that are not assessed; knowledge of the 1,000, 4,000, 6,000, 7,000, 8,000, 9,000, and 10,000+word levels is not measured. Secondly, the lists used for the different frequency levels in the VLT are rather old (Webb & Sasao, 2013). Therefore, to comprehensively assess second language learners' vocabulary size, other measures need to be employed to investigate the relationship between VB and L2 reading comprehension.
Morphological knowledge has been identified as an essential element of vocabulary depth (Nation, 2001), and therefore its relationship with L2 reading comprehension has been extensively investigated. However, research examining its role in L2 reading comprehension has produced conflicting results with some indicating its direct contribution to L2 reading comprehension, while others showing its negligible or indirect effects on L2 reading comprehension. Such discrepancy may have resulted from a lack of standardized measure of morphological knowledge as the distinct measures employed in previous studies were mostly self-made by researchers themselves, tapping different levels of morphological awareness (e.g. Qian, 1999; Zhang & Koda, 2012). Therefore, standardized measures of morphological knowledge need to be employed in future research, thus to thoroughly investigate its influence on L2 reading comprehension.
Lexical fluency or automaticity of lexical knowledge, although generally ignored in applied linguistic lexical research, has been extensively studied in the area of psycholinguistics and neurolinguistics. Therefore, it may be necessary and methodologically facilitative for applied linguists to draw on approaches from such areas
试读结束[说明：试读内容隐藏了图片]