PDF Version
The revision of the IELTS&tm; speaking testNick Boddy (British Council) |
|
Abstract |
IELTStm is a test of academic and vocational English, produced by the University of Cambridge Local Examinations Syndicate (UCLES), and jointly managed by the British Council and IDP Australia. It was launched worldwide in 1989 and is used as an indication of the readiness of examinees to participate in academic courses at institutions in many English-speaking countries.
Available at British Council offices worldwide and throughout the IDP network in the Asia Pacific region, over 100,000 took this test in 1999. Although only a small percentage of the global total take the test in Japan, many Japanese sit the test in the UK and Australia prior to embarking on higher education courses.
Apart from the UK, Australia, and New Zealand, where IELTS is the preferred entry qualification among universities, over 170 universities in the USA now welcome students with the required IELTS score. A list of institutions where IELTS is accepted is available on the Web at www.ielts.org.
While in Japan TOEFL® is now available only as a computer-based test, IELTS is only available as a pencil and paper test. A computer-based version is available elsewhere, but pilot studies conducted by the British Council in Tokyo suggested that it would not be popular among Japanese candidates. However, one drawback of using only the pencil and paper test is that the number of candidates is limited by the size of the test venue and the number of test sessions that can be offered.
The test is currently offered twice a month in Tokyo and Osaka, and less frequently in Nagoya and Fukuoka. Candidates are advised to book a minimum of four weeks in advance, as sessions are generally oversubscribed. The cost of the revised test remains unchanged at 20,000 yen.
Candidates elect to take either the Academic or the General Training modules, depending on their reason for taking the test. In Japan, around 80% opt for the Academic modules, reflecting the strong interest in studying abroad. All candidates take the same Listening and Speaking modules, but the Reading and Writing modules vary. While the Academic Reading and Writing modules assess readiness to study in English medium higher education, the General Training modules focus on the survival skills necessary in social and/or secondary education contexts. The tests are task-based, and are designed to reflect the sort of tasks candidates are likely to encounter in reality.
A band score is awarded for each module, ranging from 0, where the test was not attempted, to a maximum of 9. The global band is calculated from the average of the band scores of each module, and this is reported on a scale at 0.5 intervals. The global average for Academic module candidates in 1999 was 6.19, and among the top ten nationalities, Japan ranked 6th, with an average of 5.84. This was ahead of regional neighbours Taiwan, Thailand, and Hong Kong. The top average score by nationality was 7.0, recorded in India.
[ p. 2 ]
There is no standard required band score for entry to academic courses, as they vary greatly in terms of linguistic demands. However, very broadly speaking, a band score of around 6.0 is required for most undergraduate studies, and 6.5 to 7.0 for post-graduate studies. Some institutions also specify a minimum score in each module.
The reliability of the objectively-marked Reading and Listening tests is assured, according to UCLES, by rigorous pretesting, item analysis, and standards fixing. As a result, Cronbach Alphas of between 0.82 and 0.91 were reported for live material introduced in 1998/99 (IELTS Annual Review, 1998/1999.) Writing tasks and elicitation tasks for the Speaking test are pretested for suitability, and reliability for these tests is assured through training, certification, and monitoring of examiners. (IELTS Annual Review, 1998/1999, p. 18), details of which will be given later. However, it was recognised that in order to increase the reliability of the Speaking test, a revision of the format was necessary, and this is discussed below.
| "preparation for direct tests of writing and speaking should really involve other people; for guidance and feedback in the case of the former, and as interlocutors in the latter." |
One potential drawback for candidates used to being able to prepare for a language test by rote learning vocabulary and rules of grammar while standing on a crowded train is the fact that IELTS does not lend itself to this form of preparation.
Of course, preparation courses and materials are available, as are practice test books, and these are very valuable means of helping
candidates become familiar with the format of the test, particularly the Listening and Reading tests. However, preparation for direct
tests of writing and speaking should really involve other people; for guidance and feedback in the case of the former, and as interlocutors
in the latter. According to Cambridge, in preparing for their exams, candidates develop skills they need to make practical use of the
language, and I would argue by extension that by using the language, whether at work, academically, or socially, candidates are preparing
for the exam. For candidates remote from institutions offering preparation courses, Internet-based training for the test is already becoming
quite widely offered (see, for example,
http://international.holmesglen.vic.edu.au/IELTS01.htm [Expired Link]or
http://www.brookes.ac.uk/icels/icels_ancillary/found_acc/icelsexampreparation.html [Expired Link].
According to the 1998/99 IELTS Annual Review, discussion about the revision of the Speaking test was informed by studies conducted between 1992 and 1997. However, apart from the Merrylees and McDowell study (1999), which looked at the reliability of the Speaking test and examiner attitudes to it, and a Brown and Hill study (1998), which investigated interviewer styles and candidate performance, there appear to be no published studies directly relating to the IELTS Speaking test in general, or the Elicitation phase in particular.
This suggested that research might have been conducted by, or commissioned by UCLES/IDP for internal use, and this was confirmed by Dr. Lynda Taylor of UCLES (personal communication). Two studies were carried out on the IELTS Speaking test for UCLES by Lazaraton (1998, 2000). The first provided evidence of the effect of examiner language and behaviour on ratings, and contributed to the development of an interlocutor framework for the revised test, while the second provided confirmation of the effectiveness of the revised test, particularly the range of speech functions elicited (Taylor, personal communication). UCLES also commissioned a study of the current Speaking test (Wylie and Hudson, 1994).
[ p. 3 ]
Research into inter-rater reliability (Wylie, 1993) and intra-rater reliability of the IELTS Speaking test (Conlan, Bardsley and Martinson, 1994) was commissioned by the International Editing Committee of IELTS (Taylor, personal communication), but no details of the findings are available.
The revised Speaking test format is scheduled to be introduced worldwide in July 2001. According to UCLES, the objectives of the revision are to develop a clearer specification of tasks, in terms of input and expected candidate output; to increase standardisation of test management by the introduction of an examiner frame; and to revise the rating scale descriptors (IELTS, 2000).
The revised test will comprise three rather than five phases, but the length will remain the same. In Phase 1, which lasts 4-5 minutes, the candidate will answer general questions about themselves, their homes/families, their jobs/studies, their interests, and a range of similar familiar topic areas.
In Phase 2 (3-4 minutes) the candidate will be asked to speak for one to two minutes on a particular topic. The topic is written on a card, and the candidate has one minute to prepare the talk. S/he will be asked one or two follow-up questions.
In Phase 3 (4-5 minutes) the examiner and candidate will discuss issues and concepts thematically linked to Phase 2. An outline of the test and sample materials are available on the IELTS Web site.
Prospective IELTS Speaking and Writing examiners must first meet the IELTS Code of Practice qualification requirements. Until recently examiner training consisted of a self-access pack of a videotape, training manual, scripts to be graded, and an audiocassette of interviews to be graded. Potential examiners worked through the material, completing a number of worksheets, although these were not submitted, before awarding grades and submitting them to headquarters for checking. Grades within a certain, undisclosed range of those awarded by senior examiners resulted in examiner status being conferred for the Speaking or Writing test, or both. Failure to meet the standard resulted in the applicant being requested to reassess and resubmit the assessment.
Since 1999, centres have been encouraged to provide face-to-face training wherever possible. Two days are spent with a trainer, one on the Speaking test and the other on the Writing test. This has proved popular with trainees because it allows them to share ideas about grading, to express doubts and ask questions about the descriptors, and to engage in role-play of the Speaking test (examiner trainer and trainees, personal communication).
New examiners for the revised test will be trained after July 2001, but the first half of the year has been spent on the not inconsiderable task of retraining the cadre of current examiners worldwide. In Japan, examiner retraining has taken place in Tokyo, Nagoya, Osaka and Fukuoka, with around fifty examiners attending the one-day sessions. All examiners, current and new, will be required to grade a certification set of recordings of revised format interviews. These will be assessed in Cambridge for accuracy, and successful examiners will be issued with an identifier for use worldwide.
[ p. 4 ]
The IELTS Speaking test has been revised as a result of a number of studies which have helped to improve understanding of test and task design for assessing spoken language ability. It is to be hoped that the new test format, rigorous training, and strict monitoring of examiner performance will result in a speaking test which makes STET fairer and more reliable by candidates, teachers, examiners and receiving institutions.
Author IndexTitle IndexDate Index
Background
Links
Network
Join
[ p. 5 ]