大学英语四、六级测试现状的理论分析与问题研究

【作者】 梁佳；

【导师】 肖云南；

【作者基本信息】 湖南大学 ， 外国语言学及应用语言学， 2002， 硕士

【摘要】 语言测试发展到今天已经成为一门独立的学科，有着自己的研究领域和研究方法。语言测试从语言学、语言教学法和学习论取得科学内容，从心理测量学获得科学手段，语言测试是一门跨学科的综合性科学，语言测试又是伴随着语言教学出现的，没有语言教学也就无所谓语言测试。两者相辅相成。对与大学英语课程来说，教学最终目的是使我国大学生掌握英语，获得以英语为工具参与国际交流的能力，获取本专业所需要的各种信息；而语言测试的目的则是提供一种科学的测试工具，对学生的语言能力进行客观的、准确的、公正的评价，反映教学中的长处和短处，为提高教学质量服务。语言测试的发展经历了三个阶段：七十年代以前语言测试界占主导地位的是分析法，称为心理测量-结构主义时期，其语言学的理论根据是结构主义语学，认为语言是山语音、词汇、语法构成的一个系统，这一系统是可分的，在这一时期语言测试中使用得最广泛的题型是多项选择题：自七十年代中期起，语言测试界开始重视所谓总体综合法的研究，Spolsky等人把这一时期称为心理语言学-社会语言学时期。他们认为语言不但是一个可以分解的体系，更是一种动态的、具有创造性的功能体系。这一时期采用得比较多的是综合题，完型填空、综合改错、听写、口试、作文等等。自八十年代中期以来，随着交际教学法的发展，交际法语言测试受到了越来越多的重视。从交际法的角度来看，所谓掌握一门语言是指在一定的语境中学生能够使用所学的语言进行有效的交际，交流思想感情，达到互棚沟通的目的。不同的语言观决定语言测试不同的方法、内容和题型。一项考试的质量最主要的是体现在信度和效度两个方面。信度是对学生的语言水平提可靠的度量，效度指考试能准确地反映学生实际运用语言的能力。就语言测试这门学科目前的发展水平来看，一个考试的信度和效度常常是矛盾的。效度高的测试方式如多项选择题(指科科地设计的多项选择题)，信度很高，但效度可能不高，通过精心设计可以提高其效度，但很难达到理想的状态；效度高的测试方式，如作文、口试等，效度很高，但信度常常不高，很难保证评分的客观性和一致性，虽然经过努力，采取各种补偿方式，可以提高其信度，但同样很难达到理想的状态。这是语言测试而临的两难命题。现代语言测试尤其是大规模考试往往只能在信度和效度之间找一个最佳的平衡点，兼顾两者。交际语言测试既要考虑被试的语言形式知识，又要考虑其在有意义的语境中能 洲 人学1＠十学ft［论文 否。恰当得体他凹川这f。【心iii的能力，重点是考杏在什会涌台场景＊t。恰当有效地进行 交际的能力。传统的分o了（测试和综合测试小，被试部是被动地接受事先规定好的 测试内奔，广1己是个“周外人”。而在交际性测试小，被试变成了一个“局内人”，5 仙参与形成交际的过程。交际以六测试摆脱了分立式测试试图对综合性的语言行为 做分类式的归纳以及综合州侧试门的不明与题口之p。］汀州依赖的明显不足。分立式 测试的优点是题＊＿厅们印1－Q有独立忏，综合性测试的优点则是语境明确，交际性测 试则把两者的优点结合起来，被认为是语言测试的川Z想模式。 我网大学英语 fill、六级考试作为一干！1教学考试创搜干八－十年代小后期，是一种 尺度J：［J关一一2t｝1参照考试。j［4Jll《大学英语教学人纲》规定的教学日标考核修完 大学英语四、六级课E和＊学小nV英语水平。大学英乃课程的教学日的是培养学生具 有较狈的阅读能力和一定的听、说、读、写工 汗能力，攸学生能以英语为工具，获 1仪专业所需信息，并为进一步捉高们吾水平打－厂较奸的某汕I。这一教学日的即规定 了大学英语 卜、六级考 试的测试＊＊。门汕n大学力Z语＊、六级考试语言能力机年 模丑川。，分＿卞式讯有测试、综合币六测试以及交际旧；“Zch测试凡’i一定比例（杨惠’－【’、 “bif 1998：59－62），反吹Lll人？，地7j四、六级考试设计火对语言测试各家各派优）与 的吸收，山能窥见语jty能力观变迁扯F的痕迹。 本文在以交际法测试为川1沦杜汕＊。，对现有的大Y小“ 六级考试所采用的题型 和所存在的问题进行效但和信吐的分忻，对它对现订教’学的反拨作川进行探讨，井 设计了三下【·不同的n小任进行杆1大的问卷调查和数们统计。扮一问卷主要涉及大学英 语四、六级考仁m勺方方而而，有来门不同学校不冈午业一！。英语专业）的68O名参 加了叫、六胡考试的学牛泊‘叼了问卷。问卷内容包爪受试学生对考试的态度、看沾， 对试卷木身在时问、难度、题迎设计＿士的看沦、建议以及学生如何对待考试与语言 能力问联系的积?更多还原

【Abstract】 Language testing has gone through three stages in its development. Before the 1970s, the analytic approach known as "psychometric-structurar was in the leading position. The second stage was the popular global integrative approach. Spolsky (Spolsky, 1999) and some researchers named it the "psycholinguistic-sociolinguistic" approach. The last is the communicative approach which began in the middle 1980s. With the development of communicative language teaching, communicative language testing has been established as centrally important. According to the communicative approach, this is a method of language learning for the learner to use the language for effective communication, feeling exchanging and understanding, as well as to achieve tasks. The three approaches may seem to be divided diachronically, but they overlap each other.The emergence of the communicative era in 1970s brought a wave of criticism of the old traditional non-communicative tests. It was claimed that the type of language produced on those tests was artificial and calls were made to develop tests which would require test takers to produce REAL language, the kind of language that is used among REAL people". Consequently, in the past decade there has been a major shift in language, testing towards the development and use of communicative tests (e.g., Clark 1975, Morrow 1977; Wilds 1975). It was expected that those tests would assess a broader and more real construct of what knowing a language really means. Good tests should have sound validity and reliability, and bring advantages to teaching, known as positive backwash. The validity of a test is the extent to which it measures what it is supposed to measure and nothing else. Reliability is a necessary characteristic of any good test: for it to be valid at all, a test must first be reliable as a measuring instrumentThe CET(4/6) has been established for more than 20 years, as a national public examination, and its weak points have been revised based on the theory of communicative approach, and considering the validity and reliability of the test. In order to figure out the problems in the CET test, mainly, this paper analyzes these problems from the following three vantage points: the criterion (outline), the test type, and the backwash effect of the test. Besides, questionnaires are adopted to support conclusions in the analysis. There are three kinds of questionnaires distributed to a comparably wide range of students in different majors and teachers involved in different language-teachingcourses. As well, the paper offers some methods for their improvement. Though, the solution may be hard to carry out in application, efforts should he made.As an achievement test and a criterion-referenced test, the CET(4/6) does not reflect the actual language ability of (he students in the test completely and comprehensively. In the criterion, the description is inadequate.Today the CET(4/6) has adopted the discrete type, the integrated type and also the communicative type in testing methods in order to check the language competence. Each type has its advantages and disadvantages and that is why the CET(4/6) has combined them together to ensure a reasonable balance of validity and reliability.In the problem-analyzing part, firstly the outline is put on the table. The current situation of testees’ in the CET(4/6) js far advanced compared with those in early 90s, while the outline has had less revision. The unsuitable outlines for testing undoubtedly weaken the effect of the test. Secondly, the stereotyped items used in the GET also negatively affect the test. The uniformity of question-answer multiple-choice questions, no matter which section they are in - the listening comprehension section, the reading comprehension section or else, has made the test less authentic. Note that context is of the utmost importance in all tests. Dccontextualised multiple-choice items can do considerable harm by conveying the impression that language can be learnt and used free of any context. Both linguistic con更多还原