成绩测试中EFL写作评分标准的研究

论文摘要

写作在SLA教学和ESL测试中一直占有很大的比重,而国内外对于写作评分标准的研究大多集中在大规模考试上,对于成绩测试中能反映学生问题和影响教学的写作评分标准的研究很少。本文拟从测试方面对成绩测试中的写作评分过程、评分标准进行分析,通过考察教师在小规模的测试中对评分标准的不同理解和侧重,找出评分员之间不一致的原因,并进行有针对性地培训,最大化写作测试的评分信度,为效度验证提供前提,这也就能提高写作测试的预测效应及反拨效应,最大化成绩测试的诊断功能。本文首先回顾了国内外对于写作测试的研究,从测试方法和评分方法两方面论述了写作评分标准的使用,论述了大型考试中对于写作部分的评分细则,对于测试专家的评分标准进行了归纳总结。整体评分法由于具有较高的结构效度,能反映学生的写作能力以及高评分信度和注重学生的长处等优点,是现在最常使用的评分方法。但为了实现评分信度和写作测试的效度,就需要制定合适的评分标准来规范评分过程,而目前在成绩测试中使用的写作评分标准大多是直接照搬大型考试的评分标准,在内容效度、真实性以及可行性上都有不太适合成绩测试大纲规定的地方。本研究随机抽取了10份大三学生期末考试写作试题作为样本,邀请10位教师作为评分员,要求他们不参照任何评分标准用整体评分法进行评分,列出3条评分依据并写出详细的评分思路。本文运用SPSS对写作评分成绩进行了信度、效度的定量分析;参照Ling Shi(2001)的编码方法归纳了评分员给出的评分依据,并对评分员的评分过程进行了定性分析。用归纳的写作评分标准再次进行评分,并对评分结果进行了信、效度分析。研究结果表明,对评分结果用相关性系数以及T检验没有发现显著差异,评分员用整体评分法所给出的分数接近但又能区分出样本等级;通过对评分员评分过程的分析,我们发现评分员采用了相似的评分标准,都是从内容、结构和语言的准确性给分,但主观性也很大,对样本所反映的语言能力高低有不同的认识,应加强对评分员的培训以最大化信、效度;由于测试类型和测试内容的不同,成绩测试的写作评分应考虑教学大纲中对于语言能力的要求,以区别于大型考试的评分标准。

论文目录

• Table of Contents in English
• Table of Contents in Chinese
• List of Tables
• Abstract
• 中文摘要
• Chapter Ⅰ Introduction
• 1.1 Reseach background
• 1.2 Purpose and significance of the study
• 1.3 Major concepts
• 1.4 Research questions
• 1.5 Outline of the paper
• Chapter Ⅱ Literature Review
• 2.1 Overview of ESL writing assessment
• 2.1.1 Two distinct methods for assessing writing ability
• 2.1.1.1 Direct writing
• 2.1.1.2 Indirect writing
• 2.2 Scoring methods in writing assessment
• 2.2.1 Mechanical method
• 2.2.2 Holistic scoring
• 2.2.3 Primary trait scoring
• 2.2.4 Analytic scoring
• 2.3 Empirically-based rating criteria of writing in large-scale exams
• 2.3.1 Rating criteria of writing by holistic scoring
• 2.3.1.1 ILR's criteria of writing assessment
• 2.3.1.2 Writing assessment for TEM 4 and 8
• 2.3.1.3 Rating criteria of writing in CET 4 and 6
• 2.3.2 Rating criteria of writing by analytic scoring
• 2.3.2.1 The R.S.A.C.R.E.F.L.Examination
• 2.3.2.2 IELTS Criteria of Assessing writing
• 2.4 Empirically-based rating criteria of writing by researchers
• 2.4.1 Lado's evaluation of content and style
• 2.4.2 Oller's method of rating content and organization
• 2.4.3 20 characteristics summarized by Breland & Jones
• 2.4.4 Weir's TEEP experience
• 2.4.5 Sasaki and Hirose's development of rating scale
• Chapter Ⅲ Methodology
• 3.1 Research design
• 3.1.1 Teacher raters
• 3.1.2 Written samples
• 3.1.3 Methodology
• 3.2 Procedures
• 3.2.1 Scoring without any criteria
• 3.2.2 Reeoding the key elements and retrospective interview
• 3.2.3 Rater training and pilot testing of the obtained rating criteria
• 3.2.4 Reliablility and validity of the study
• 3.2.4.1 Reliability of the present study
• 3.2.4.2 Validity of the present study
• 3.3 Summary
• Chapter Ⅳ Results and Discussion
• 4.1 Analysis of the first rating results
• 4.1.1 Quantitative analysis on the test scores
• 4.1.1.1 Descriptive statistics
• 4.1.1.2 Reliability analysis of the scores
• 4.2 Analysis of the raters' reports
• 4.3 Analysis of the rating criteria given by the raters
• 4.4 Rerating and pilot testing the criteria
• 4.4.1 Procedure
• 4.4.2 Quantitative analysis of the re-rating scores
• 4.4.2.1 Descriptive analysis
• 4.4.2.2 Reliability analysis
• 4.4.2.3 Validity analysis
• 4.5 Summary
• Chapter Ⅴ Conclusion
• 5.1 Findings and implications of the research
• 5.2 Limitations of the current research
• 5.3 Recommendations for future research
• Bibliography
• Appendix
• Appendix 1 Sample 2
• Appendix 2 Questionarire
• Appendix 3 Rating Report Card l
• 发表文章目录
• Acknowledgements
• 个人简况及联系方式