作者：吴越
出版社：上海交通大学出版社
格式: AZW3, DOCX, EPUB, MOBI, PDF, TXT
EFL综合性写作测试任务受试策略运用研究(英文版)试读：
前言
语言测试的本质在于根据受试完成目标语言任务的情况，对其语言交际能力做出准确推断及合理解释（Bachman＆Palmer，2010），因而，对语言测试任务及其所测构念（construct）加以描述并找到两者之间的匹配关系是开发测试任务的重要基础。在整体语言观的指导下，融合多种语言技能的综合型测试任务（integrated tasks）已成为21世纪语言测试的显著特征之一（金艳，2013：3）。综合性写作测试在学术英语写作语境中具有独特优势。首先，这类测试任务很大程度上模拟、复制了学术写作语境对受试写作能力的要求，能够较好地预测受试者在真实学术语境中的写作能力，因而具有较高的真实性（authenticity）（Lewkowicz，1997；Plakans，2009）和预测效度（predictive validity）（Bachman，1990）。再者，学术语境中的写作任务通常不可能单独存在，学习者需要理解、分析、概括、阐释、整合来自阅读材料的信息，并在此基础之上完成写作（Weigle，2004：34），综合性写作测试通过为全体受试提供同质同量的语言输入，避免了写作话题可能带来的内容偏颇（bias），具有较高的考试公平性（test fairness）（Weir，2010）。众多国内外大规模高风险考试采用了综合性写作测试任务替代传统写作测试项目，如托福网考（TOEFLiBT），台湾地区全民英语能力分级检定考试（GEPT）、大学英语四、六级网考和高考英语广东卷等，这种融合多种技能的综合性写作评测模式已成为语言教学和语言测试领域中一项新兴的重要研究课题。然而，以往研究过多关注受试写作文本特征分析或阅读与写作独立技能之间的关系，而鲜有研究者对EFL学习者在读写过程中究竟如何运用策略做过系统性研究。
基于文献回顾，本书提出综合性写作测试中受试写作策略的构成假设框架，其中包括：意义构建策略（meaning-constructing strategy use）、元认知策略（metacognitive strategy use）和认知策略（cognitive strategy use）。根据此假设，本书旨在回答以下问题：第一，综合性写作任务中受试策略运用这一构念的本质是什么？换言之，任务完成过程中，受试究竟运用了哪些策略并如何使用这些策略？第二，不同水平组受试在策略运用方面有无差异？如果有，体现在哪些方面？
本书以读写结合写作任务为测试工具（该测试项目效度已在相关文献中得以验证），收集以下数据：①约212名受试综合性写作文本；②3位评分员分别对212篇作文的评分数据，获得有关评分质量和评分维度的信息为深入分析提供背景信息；③关于读写结合综合性写作测试策略运用因素的问卷，探究受试写作策略运用构成因素，并验证模型的合理性；④16位受试有声思维和回溯性访谈录音数据，旨在对受试策略运用做出推断。
本书从四个方面对所收集数据进行了分析：①利用多层面Rasch模型分析写作任务分数数据；②利用SPSS 17.0对问卷数据进行探索性因子分析；③利用AMOS 17.0进行验证性因子分析；④利用NVivo 8.0对有声思维和回溯性访谈录音进行转写、编码，并进行定量统计和定性分析。
本书得出以下结论：①受试在任务完成过程涉及先验知识与原文相互作用中所建构的三种策略构念之间的互动，其中意义构建策略运用包括选择、组织和连接策略；元认知策略运用包括整体构思、局部构思、宏观监察、微观监察、读者意识和评价策略；认知策略运用包括推断、阐明和信息管理策略。②不同水平组受试策略运用存在差异，主要体现在：高分组受试能够较为有效地运用选择、组织、连接、整体构思、宏观监察、读者意识、评价、阐明和信息管理策略；低分组则主要集中运用局部构思、微观监察和推断策略。
本书对于综合性写作任务设计与开发、策略运用构念研究、综合性写作任务评分标准的选择和制定、综合性写作教学具有启示意义。
本书于2014年获上海市教育科学研究项目（B14042）资助，谨表谢忱。CHAPTER ONEINTRODUCTION
ESL learners’integrated writing ability has been increasingly considered important for academic purposes based on the assumption that writing should be associated with what learners are supposed to do in academic learning contexts in the sense that they are often expected to comprehend，evaluate，and analyze source texts for the purpose of both generating and synthesizing ideas into written products（Cumming 2013；Weigle et al.2013；Cumming&Silva 2008）.For academic purposes，thus，the writing ability construct should comprise the ability not only to generate coherent written products as measured by the conventional writing-only task type but also to compose texts that are purposefully and appropriately relevant to ideas and information from academic sources（Plakans 2009；Cumming 2013）；however，there is scant empirical evidence about how learners use a repertoire of strategies while performing those tasks.Findings from research on this task type can not only provide evidence that helps understand the extent to which subjects’strategy use interacts with their reading-to-write test performance but offer insights into ESL/EFL writing instruction.With these aims in mind，this research attempts to investigate subjects’strategy use in completing an integrated reading-to-write task and to what extent their deployment of strategies differs across scoring levels.1.1 Rationale for the study
Second language writing research is characterized by its complexity and multidisciplinary nature（Matsuda，2006）.Most researchers acknowledge that writing can be essentially described as“a‘more’and‘better’incorporation of various separate attributes that make up language performance”（Byrnes 2002：37）.In particular，the communicative competence framework，established by Hyland（2003），highlighting that ESL learners require at least grammatical competence，discourse competence，sociolinguistic competence，and strategic competence for the purpose of writing successfully in English.This framework“sheds light on the varying needs of students as they progress from Novice to Distinguished levels of proficiency”（Leaver&Shekhtman 2002：10）.The development of learners’communicative competence has long been perceived as the fundamental objective of ESL/EFL teaching.In achieving such a goal，the integrated-skill approach has increasingly been favored by both language instructors and assessors because of its simulation of authentic language use in real life，its stimulation of learners’strategy use，and its encouragement of learners’acquisition of content knowledge by means of multiple methods（Grabe&Stoller 2001）.
With regard to instruction，language instructors have attempted to simulate authentic language use in classroom settings in order to help learners better prepare for language tasks in real-life settings.Specifically，when learners are requested to perform complicated tasks in academic contexts，language instructors have to develop their integrated language skills by implementing instructions in which different language components are combined.
Similar benefits are applicable to performance assessment.In writing assessment，Weigle（2004）claimed that writing is seldom done in isolation，but is essentially conducted as a response to source texts instead.In recent years，furthermore，conventional impromptu writing-only tasks have been criticized as an inauthentic task type that cannot provide a valid assessment of academic writing ability in real language contexts（Hamp-Lyons&Kroll 1996；Leki&Carson 1997；Weigle 2002，2004）.As a response to this long-standing criticism，the use of integrated writing tasks for assessing writing ability in large-scale language assessment programs has been increasing in recent years（e.g.，General English Proficiency Test[GEPT]，Canadian Academic English Language Assessment[CAEL]，Internet-based Test of English as a Foreign Language[TOEFL iBT]，Certificate of Proficiency in English[COPE]），often as a substitute for impromptu writing-only tasks.The primary rationale for the integrated writing task type consists in the general assumption that writing test tasks for academic purposes should be associated with what learners are supposed to do in real language contexts in the sense that students in colleges or universities are frequently expected to write for the purpose of demonstrating their writing proficiency in different skill areas，including listening or reading（Cumming 2013；Weigle et al.2013；Leki，Cumming，&Silva 2008）；therefore，for academic purposes，the writing ability construct should comprise the ability not only to generate coherent written texts as measured by the conventional writing-only task type but also to compose written responses that are purposefully and appropriately related to ideas and information from given academic sources（Plakans 2009；Cumming 2013）.Research interest has accordingly been growing，as proved by a body of studies on integrated test tasks in recent years（e.g.，Cumming et al.2005；Asención 2004；Gebril 2009；Plakans 2008，2009；Yang&Plakans 2012；Kim 2013）.The intense research interest in the promising task type could also be evidenced by the fact that one of the most notable academic periodical in the field of language testing，Language Assessment Quarterly，published a special issue in 2013，addressing both theoretical rationales and empirical evidence for integrated writing test tasks.
Additionally，the past decades have witnessed the shift of research paradigm from how second language or foreign language should be taught to what makes a good language leaner，which has promoted researchers’interest in language learning strategies invoked by ESL/EFL learners and their effects on the development of language proficiency（e.g.，O’Malley&Chamot 1987；Wenden 1987；Abraham&Vann 1987；Purpura 1999）.Just a few researchers（e.g.，Tsai 2004；Ou 2013），however，have systematically concentrated on the writing strategies of good writers or what strategies help writers develop their writing skills.As claimed by both Victori（1995）and Beare（2000），it is the strategy use that primarily distinguishes successful from less successful writers，which has also been supported by Sasaki（2000）who found that expert and novice writers made use of different strategies or used the same ones but in different ways.Furthermore，it has been found that test-taking strategy research provides insights for test validation.In order to achieve a better understanding of the construct of test tasks，language practitioners should，as Bachman（2002）suggested，
collect data on the processes or strategies that test-takers use in responding to assessment tasks（e.g.，verbal protocols，observations，questionnaires，interviews，discourse that is created in the assessment process，such as speech or writing samples，physiological and neurophysiological responses）and utilize appropriate qualitative analyses to investigate the ways in which test-takers process language assessment tasks（p.470）.
In this sense，many areas concerning integrated writing tasks remain to be more fully explored.In particular，there is insufficient empirical evidence about how subjects use different strategies while performing those tasks.Further exploration is needed in terms of the extent to which subjects’strategy use interacts with their reading-to-write test performance.
Three research issues underlie the present study：（1）a lack of empirical research on strategy use reported by subjects while performing integrated writing tasks in Chinese EFL contexts；（2）a lack of information concerning the interrelatedness between EFL university students’strategy use and their performance on integrated writing tasks across scoring levels；（3）the scarcity of research employing a mixed-method approach for investigating strategy use.
First，it is noteworthy that there has been a lack of empirical study focusing specifically on subjects’reported strategy use while performing an integrated reading-to-write task particularly in Chinese EFL contexts.While numerous previous studies explored how language learners employ strategies in an attempt to acquire a language skill，very few focused on their strategy use in assessment contexts（Tsai 2004；Yang 2009；Yang&Plakans 2012）.Additionally，according to Macaro（2007），learners’combined uses of a strategy with others could improve performance in a given task.There is，thus，an obvious lack of systematic research addressing the issue of how EFL subjects actually deploy meaning-constructing，metacognitive and cognitive strategies when engaged in an integrated reading-to-write task.
Second，much in-depth empirical information is needed to reveal the interrelatedness between EFL university students’strategy use and their performance on integrated writing tasks across scoring levels.It has been believed that the use of strategies（e.g.，metacognitive type）awakens learners’awareness to recognize learning contexts and processes and thus leads to more efficient learning and improved performance（Anderson 2002）.Thus，we essentially need more substantial evidence to address the complexities of the relationship between subjects’strategy use and their performance.
Third，concerning research methodology，most of the previous studies have adopted a single-method approach：either quantitative（e.g.，Cumming et al.2005；Yang&Plakans 2012）or qualitative research method（e.g.，Delaney 2008；Plakans 2006，2008）.There is a need for explicit and systematic use of mixed-method designs to concurrently collect different but complementary evidence which allows“triangulation”to investigate subjects’strategy use corresponding to interactive processes of writing from sources.
The purpose of this study is to explore the nature of subjects’strategy use construct in their completion of an EFL integrated reading-to-write task as represented by its generic components on the basis of the evidence from both questionnaires and qualitative think-aloud protocol（TAP）data.It also aims at specifying differences in terms of strategy use across scoring levels as revealed in TAPs and follow-up interviews.
With regard to data，the study involved four types of data：1）integrated writing test performance derived from 212 subjects；2）a questionnaire focusing on strategy use issued to these subjects immediately after the integrated writing test in attempt to elicit factors contributing to the construct of subjects’strategy use；3）16 subjects’TAPs while performing the integrated writing task recorded，transcribed and coded to derive information from which an inference concerning their strategy use could be made；4）follow-up interviews to the 16 TAP participants conducted with the intention of exploring their utilization of a variety of strategies.
The data were analyzed from the following four perspectives：1）Multi-facets Rasch Model（MFRM）was used to analyze the facets of the test that may contribute to variability of the test scores；2）exploratory factor analyses were conducted to identify variables from questionnaire data with which generate initial models of strategy use；3）structural equation modeling was utilized to measure the quality of these models.4）16 participants’TAP reports and follow-up interview transcripts were coded and qualitatively analyzed by means of Nvivo 8.0.The qualitative data elicited from TAPs and interviews were used to triangulate quantitative findings and provide further information about the strategy use of subjects across groups of different scoring levels.1.2 Significance of the study
Given the present study investigates，both qualitatively and quantitatively，the nature of subjects’strategy use and their performance on an integrated reading-to-write task，it has theoretical，pedagogical and methodological implications.
First，by examining the nature of strategy use and integrated writing test performance through quantitative and qualitative lenses，the study proposes a model that reveals subjects’interactive deployment of strategies in their task fulfillment，which may definitely expand our understanding of how they approach integrated writing tasks.A body of studies in the field of integrated language test tasks have delved into either subjects’written products（e.g.，Gebril&Plakans 2013；Watanabe 2001；Cumming et al.2005）or their performing processes（e.g.，Plakans 2007，2008，2009；Plakans&Gebril 2012；McCulloch 2013）.Still other studies sought to justify the validation of this task type（Asención 2004；Delaney 2008；Kim 2008；Jiang 2010）.Little research，however，has unearthed the generic components of subjects’strategy use in their completion of an EFL integrated reading-to-write task，in particular，meaning-constructing strategies，cognitive strategies，and metacognitive strategies.
Second，since this study investigates to what extent the strategy use that subjects report may influence their performance on reading-to-write tasks，it could have pedagogical implications on EFL writing.In other words，if it is found that the specific strategy use，i.e.，meaning-constructing，cognitive and metacognitive strategies，have differential effects on subjects’performance，this might shed light on how EFL learners should be taught in integrated writing particularly in academic contexts.As Keck（2006）suggested，writing instructors present learners with examples of both good and poor writing strategy use in order to help them differentiate between acceptable and unacceptable writing practices.Furthermore，integration of multiple skills makes this task type a different assessment of writing which requires subjects to count on a wide repertoire of strategies and to make challenging decisions.Therefore，writing classes should prioritize these strategies in instruction and practice.
In addition，as this study involves the construction and validation of strategy use assessing instruments with regard to reading-to-write tasks primarily on the basis of previous research（Yang 2009；Tsai 2004；Purpura 1997；Plakans 2009），EFL teachers could be provided with a taxonomy of strategy use for academic writing because these instruments could offer useful diagnostic information about learners’reading-to-write performance and their strategy use so that teachers could train their learners to utilize in an attempt to improve their academic writing abilities.1.3 Outline of the study
The remaining part of the dissertation is made up of the following five chapters：
Chapter Two first reviews Constructivist discourse synthesis theory，theoretical concepts of both cognitive and metacognitive strategy use in writing processes，and previous empirical studies on integrated writing tasks.Upon the review of literature，a hypothesis concerning the construct of subjects’strategy use in integrated writing tasks is presented，and research questions are indicated.
Chapter Three describes the methodology taken in the present research，including data collection methods，i.e.，questionnaires，TAPs，and follow-up interviews，quantitative data analysis methods，i.e.，Factor analysis，Multi-facets Rasch Model and qualitative data analysis methods by means of Nvivo 8.0.This chapter aims to build a coherent link between the rationale of data triangulation and selection of research methods.
Chapter Four depicts the results of both quantitative and qualitative data.First，the MFRM analysis is presented to reveal facets that might contribute to variability of the test scores.Questionnaire survey results will be indicated，including exploratory factor analysis and confirmatory factor analysis using SEM，to address the nature of subjects’strategy use.Subjects’TAP and interview data are analyzed thereafter with due attention to the differences of strategy use across scoring levels.
Chapter Five steps further into the discussion of the research findings within the broader context of the fields of language assessment and educational psychology.Each of the research questions will be discussed first in light of the research results presented in the previous chapter.This discussion will be related to the previous studies concerning mental operations in discourse synthesis as well as other process-oriented research on integrated writing test tasks.At the end of this chapter，the view of the present research on the nature of subjects’strategy use is proposed.
Chapter Six makes a brief conclusion，addresses the limitations of the present research and the directions for future research，and presents the theoretical，methodological，and pedagogical implications of the research for second language writing and testing research.CHAPTER TWOLITERATURE REVIEW
Most educators have accepted the assumption that learning strategy use serves as an indicator differentiating low-performing from high-performing learners（Brown et al.1983）.Given this assumption，earlier research in the field of second language acquisition（SLA）has also begun to recognize the effects of learning strategies on learners’language acquisition（e.g.，Cohen 1984，1987；Clahsen 1987；Chamot 1987；Wenden 1991），a few of which found that these strategies are associated with performance（e.g.，Politzer and McGroarty 1985；Huang and Van Naerssen 1985；Jamieson&Chapelle 1987）.Since one important aspect of language assessment research is the investigation into the factors that account for variance in test scores，the findings of the aforementioned strategy studies have begun to arouse interest from testers.As Purpura（1996）figured out，
“...as strategy use may very well turn out to have a significant impact on the variation of test scores.For this reason，the investigation into the relationships between strategy use and language test performance is an essential part of the process of construct validation of language tests with far-reaching implications for language educators and testers alike”（p.23）.
This chapter synthesizes relevant theoretical and empirical research that has contributed to the rationale of the present study.The focus of this research is subjects’strategy use and their performance on an integrated reading-to-write task.It attempts，through investigating subjects’performing processes，to establish a framework that is supposed to be constructed by the contribution of various components of strategy use and，at the same time，to describe the nature of the strategy use.Therefore，it is essential to know，firstly，the mental operations in writing since those discussions could be informative concerning what writers actually proceed in mind and what specific strategies they draw on while writing an essay as a response to a certain given topic，which constitutes the first pillar for the review of literature.The second pillar comprises the constructivist discourse synthesis theory since what makes reading-to-write construct different from that of a writing-only task lies in the providing of source texts，and reading-to-write behaviors in this study are perceived as meaning-constructing processes highly dependent on the writers’ability to synthesize information from source texts；therefore，the theory elaborating organizing，selecting and connecting content from given source texts provides insights into the study of strategy use in integrated writing tasks.
试读结束[说明：试读内容隐藏了图片]