论文部分内容阅读
【Abstract】 Testing the ability to speak is a most important aspect of language testing. Testing of students’ spoken English can bring in a positive washback for English teaching. However， it’s extremely difficult to test. This paper conducts a general analysis of oral production tests from overview， types and items of oral tests and scoring methods， which’s supposed to serve as an introduction to the issue under discussion.
【Key Words】 Spoken English ability； Oral production tests； Scoring methods
【中圖分类号】G642.05 【文献标识码】A 【文章编号】2095-3089（2013）33-0-02
1. Overview of oral production tests
Oral production testing can examine and evaluate students’ competence of using oral English with high validity. But the assessments for students’ spoken English are mainly from scorers’ subjective judgment， which makes it with low reliability. Before we study deeper into oral production tests， we should have a thorough understanding.
1.1 Nature of the spoken language
According to Tepper （1978）， oral communication activity can be defined as “an activity between two or more people within which people communicate and accept ideas， feelings and attitude through verbal and nonverbal methods”. Therefore the oral communication activity has following features： interactiveness， spontaneity， purposefulness， para-linguistic features， non-linguistic features， inseparability of listening from speaking.
Which specific aspects can reflect a person’s oral proficiency？ Weir & Bygate （1992） pointed that oral proficiency was formed by three parts： routine skills （information/interaction routines）， improvisation skills （ability to tactically resolve problems encountered in the process of communication） and micro-linguistic skills （knowledge of phonology， grammar， and vocabulary）.These three skills represent three levels of oral proficiency， working together to make the communication successfully.
1.2 Bachman’s model of oral test performance
Bachman （2001） proposed a model of oral test performance. In the model there are various variables involved in oral assessment such as rater， candidate， speech sample， scale criteria， etc. All these variables formed a tight web. They may influence each other， or one variable is influenced by others and then gains influence on another. It fully reflects the complexity of oral assessment-so many variables influence each other in a dynamic way， which on the one hand， helps us better understand the potential difficulties in assessing oral language， and on the other hand reminds us that all the factors should be taken into consideration while designing oral assessment. 1.3 Difficulties in testing speaking skills
Testing the ability to speak is an extremely difficult skill to test as it is too complex a skill to permit any reliable analysis to be made for the purpose of objective testing which is of high validity but low reliability.
1） Questions relating to the criteria for measuring the speaking skills remain unanswered. 2） Impossibility of separating the speaking skills from the listening skills. 3） Administration， impossible to test large numbers of students because of limited time involved. 4） The discrepancy of scorers’ professional skills， testing experiences and personality inevitably influence their judgments， which makes it difficult to reach an absolute agreement in grasping the evaluation standard.
All these factors make accurate measurement of oral ability not easy. It takes considerable time and effort to obtain valid and reliable results.
2. Types and items of oral tests
Although oral production test of foreign languages has undergone a comparatively short history， yet it has attracted more and more attention in the recent years. In following part， we’re going to study oral production test from two aspects： types and item writing.
2.1 Types of oral tests
According to Clark （1979）， there are three oral assessment methods： direct， indirect and semi-direct. Direct oral assessment is “procedures in which the examinee is asked to engage face-to-face communicative exchange with one or more human interlocutors”， including oral interviews， pair interaction and group oral. Peer-to-peer interaction has emerged as an alternative format to the traditional oral interviews， producing more varied， balanced， and authentic discourse.
Indirect oral assessment generally refers to assessing speaking not directly， which means the oral tasks which do not require the test-taker to speak. This kind of method is not suitable for communicative era and hence almost abandoned.
Semi-direct oral assessment employs “the means of tape recordings， printed test booklets， or other non-human elicitation procedures rather than face-to-face conversation with a live interlocutor” （Clark， 1979）. It usually takes place in the language laboratory where several candidates respond to pre-recorded materials through earphones.
2.2 Common items of oral tests
Bachman （1990） classified items of oral production test into two categories： selected response items and constructed response items. The former focuses on testing receptive skills， micro-linguistic skills and part of routine skills. Besides， it’s also indirect oral assessment. It’s not suitable for oral production tests. The latter can be subdivided into non-communicative items and communicative items due to how many examinees engage in the communication （Zou， 2005）. In this paper， several common items—reading aloud， picture description， oral interview and role play will be discussed in detail.
Reading aloud： Student is given a short time to glance through an extract before being required to read it aloud. Tests involving reading aloud are generally used to assess pronunciation as distinct from the total speaking skills. The backwash effects may be very harmful. In order to construct suitable tests of reading aloud， it’s helpful to imagine actual situations in real life in which the testees may be required to read aloud.
Picture description： Pictures， maps and diagrams can be used in oral production tests. Pictures of single objects can be used for testing the production of significant phoneme contrasts， while a picture of a scene or an incident can be used for examining the total oral skills.
Oral interview： “the most ideal form of oral communication testing”. It offers a realistic means of assessing the total oral skill in a “natural” speech situation. However， the scoring is highly objective and performance in a particular interview may not accurately reflect testee’s true ability because of tensions etc. To ensure the validity and reliability of oral interview test， four points should be assured： interview should begin with easier questions； two examiners； forming a pair with similar personalities and levels of language ability； task should be helpful.
Role play activities can also be used successfully to test oral communicative ability. The students involved are assigned fictitious roles and are required to improvise in language and behavior. It is advisable for the students to be given fictitious names before the role play， and a card on which there are a few sentences describing what kind of a person he or she is， etc.
3. Scoring methods
There are two widely used scoring methods： holistic and analytic scoring. Holistic scoring is based on the global impression of the candidate’s performance and represents the rater’s overall judgment and lays emphasis on the function and effectiveness of language communication. Analytic scoring is a fairly detailed marking scheme， showing accuracy of pronunciation， grammar， vocabulary， appropriacy， fluency， and ease of speech， focusing on the accuracy of language structure and grammar rules （Wen， 1999）.
Heated debates have been aroused on which scoring method is better. Holistic scoring has potential problems： it provides very little information of the candidate’s actual performance and cause improper interpretations of the score. By contrast， analytic scores can provide diagnostic information for examinees with varied profiles. The main problem is that it may cause high cognitive load on raters as well as difficulty in defining the dimensions in analytic rubrics precisely. Therefore， certain suggestions are provided to make a valid and reliable scoring： 1） Clearly recognizable and appropriate descriptions of criterial levels are written and scorers are trained to use them. 2） Irrelevant features of performance are ignored. 3） There is more than one scorer for each performance.
4. Conclusion
The accurate measurement of oral ability is not easy. The study in the field of testing spoken language has only a short history； it can not compete against the study on other conventional tests such as writing， reading， listening， etc. The present review then is nothing but a tentative attempt against such an academic background. All that has been discussed above is supposed to serve as an introduction to the issue under discussion so others might join and come up with more valuable and insightful views about it.
Biography
[1] Bachmam， L. F. 1990. Fundamental Considerations in Language Testing [M]. Oxford： Oxford University Press.
[2] Clark， J. L. D. 1979. An exploration of speaking proficiency measures in the TOEFL context. Princeton， NJ： Educational Testing Service.
[3] Tepper， A. 1978. Speech Communication Theory [M]. Kendal： Hunt Publishing Company.
[4] Weir， C. J. & Bygate， M. 1992. Meeting the criteria of communicativeness in a spoken language test [J]. Journal of English and Foreign Languages.
[5]文秋芳，1999，《英語口语测试与教学》[M].上海：上海外语教育出版社.
[6]邹申，2005，《语言测试》[M].上海：上海外语教育出版社.
【Key Words】 Spoken English ability； Oral production tests； Scoring methods
【中圖分类号】G642.05 【文献标识码】A 【文章编号】2095-3089（2013）33-0-02
1. Overview of oral production tests
Oral production testing can examine and evaluate students’ competence of using oral English with high validity. But the assessments for students’ spoken English are mainly from scorers’ subjective judgment， which makes it with low reliability. Before we study deeper into oral production tests， we should have a thorough understanding.
1.1 Nature of the spoken language
According to Tepper （1978）， oral communication activity can be defined as “an activity between two or more people within which people communicate and accept ideas， feelings and attitude through verbal and nonverbal methods”. Therefore the oral communication activity has following features： interactiveness， spontaneity， purposefulness， para-linguistic features， non-linguistic features， inseparability of listening from speaking.
Which specific aspects can reflect a person’s oral proficiency？ Weir & Bygate （1992） pointed that oral proficiency was formed by three parts： routine skills （information/interaction routines）， improvisation skills （ability to tactically resolve problems encountered in the process of communication） and micro-linguistic skills （knowledge of phonology， grammar， and vocabulary）.These three skills represent three levels of oral proficiency， working together to make the communication successfully.
1.2 Bachman’s model of oral test performance
Bachman （2001） proposed a model of oral test performance. In the model there are various variables involved in oral assessment such as rater， candidate， speech sample， scale criteria， etc. All these variables formed a tight web. They may influence each other， or one variable is influenced by others and then gains influence on another. It fully reflects the complexity of oral assessment-so many variables influence each other in a dynamic way， which on the one hand， helps us better understand the potential difficulties in assessing oral language， and on the other hand reminds us that all the factors should be taken into consideration while designing oral assessment. 1.3 Difficulties in testing speaking skills
Testing the ability to speak is an extremely difficult skill to test as it is too complex a skill to permit any reliable analysis to be made for the purpose of objective testing which is of high validity but low reliability.
1） Questions relating to the criteria for measuring the speaking skills remain unanswered. 2） Impossibility of separating the speaking skills from the listening skills. 3） Administration， impossible to test large numbers of students because of limited time involved. 4） The discrepancy of scorers’ professional skills， testing experiences and personality inevitably influence their judgments， which makes it difficult to reach an absolute agreement in grasping the evaluation standard.
All these factors make accurate measurement of oral ability not easy. It takes considerable time and effort to obtain valid and reliable results.
2. Types and items of oral tests
Although oral production test of foreign languages has undergone a comparatively short history， yet it has attracted more and more attention in the recent years. In following part， we’re going to study oral production test from two aspects： types and item writing.
2.1 Types of oral tests
According to Clark （1979）， there are three oral assessment methods： direct， indirect and semi-direct. Direct oral assessment is “procedures in which the examinee is asked to engage face-to-face communicative exchange with one or more human interlocutors”， including oral interviews， pair interaction and group oral. Peer-to-peer interaction has emerged as an alternative format to the traditional oral interviews， producing more varied， balanced， and authentic discourse.
Indirect oral assessment generally refers to assessing speaking not directly， which means the oral tasks which do not require the test-taker to speak. This kind of method is not suitable for communicative era and hence almost abandoned.
Semi-direct oral assessment employs “the means of tape recordings， printed test booklets， or other non-human elicitation procedures rather than face-to-face conversation with a live interlocutor” （Clark， 1979）. It usually takes place in the language laboratory where several candidates respond to pre-recorded materials through earphones.
2.2 Common items of oral tests
Bachman （1990） classified items of oral production test into two categories： selected response items and constructed response items. The former focuses on testing receptive skills， micro-linguistic skills and part of routine skills. Besides， it’s also indirect oral assessment. It’s not suitable for oral production tests. The latter can be subdivided into non-communicative items and communicative items due to how many examinees engage in the communication （Zou， 2005）. In this paper， several common items—reading aloud， picture description， oral interview and role play will be discussed in detail.
Reading aloud： Student is given a short time to glance through an extract before being required to read it aloud. Tests involving reading aloud are generally used to assess pronunciation as distinct from the total speaking skills. The backwash effects may be very harmful. In order to construct suitable tests of reading aloud， it’s helpful to imagine actual situations in real life in which the testees may be required to read aloud.
Picture description： Pictures， maps and diagrams can be used in oral production tests. Pictures of single objects can be used for testing the production of significant phoneme contrasts， while a picture of a scene or an incident can be used for examining the total oral skills.
Oral interview： “the most ideal form of oral communication testing”. It offers a realistic means of assessing the total oral skill in a “natural” speech situation. However， the scoring is highly objective and performance in a particular interview may not accurately reflect testee’s true ability because of tensions etc. To ensure the validity and reliability of oral interview test， four points should be assured： interview should begin with easier questions； two examiners； forming a pair with similar personalities and levels of language ability； task should be helpful.
Role play activities can also be used successfully to test oral communicative ability. The students involved are assigned fictitious roles and are required to improvise in language and behavior. It is advisable for the students to be given fictitious names before the role play， and a card on which there are a few sentences describing what kind of a person he or she is， etc.
3. Scoring methods
There are two widely used scoring methods： holistic and analytic scoring. Holistic scoring is based on the global impression of the candidate’s performance and represents the rater’s overall judgment and lays emphasis on the function and effectiveness of language communication. Analytic scoring is a fairly detailed marking scheme， showing accuracy of pronunciation， grammar， vocabulary， appropriacy， fluency， and ease of speech， focusing on the accuracy of language structure and grammar rules （Wen， 1999）.
Heated debates have been aroused on which scoring method is better. Holistic scoring has potential problems： it provides very little information of the candidate’s actual performance and cause improper interpretations of the score. By contrast， analytic scores can provide diagnostic information for examinees with varied profiles. The main problem is that it may cause high cognitive load on raters as well as difficulty in defining the dimensions in analytic rubrics precisely. Therefore， certain suggestions are provided to make a valid and reliable scoring： 1） Clearly recognizable and appropriate descriptions of criterial levels are written and scorers are trained to use them. 2） Irrelevant features of performance are ignored. 3） There is more than one scorer for each performance.
4. Conclusion
The accurate measurement of oral ability is not easy. The study in the field of testing spoken language has only a short history； it can not compete against the study on other conventional tests such as writing， reading， listening， etc. The present review then is nothing but a tentative attempt against such an academic background. All that has been discussed above is supposed to serve as an introduction to the issue under discussion so others might join and come up with more valuable and insightful views about it.
Biography
[1] Bachmam， L. F. 1990. Fundamental Considerations in Language Testing [M]. Oxford： Oxford University Press.
[2] Clark， J. L. D. 1979. An exploration of speaking proficiency measures in the TOEFL context. Princeton， NJ： Educational Testing Service.
[3] Tepper， A. 1978. Speech Communication Theory [M]. Kendal： Hunt Publishing Company.
[4] Weir， C. J. & Bygate， M. 1992. Meeting the criteria of communicativeness in a spoken language test [J]. Journal of English and Foreign Languages.
[5]文秋芳，1999，《英語口语测试与教学》[M].上海：上海外语教育出版社.
[6]邹申，2005，《语言测试》[M].上海：上海外语教育出版社.