Chapter 1 lays the theoretical foundations. A brief historical introduction sets the scene, giving examples of
language learning for business use, and an overview of business language tests. The author then summarizes the
theory of business language testing, suggesting that business language testing is a type of testing language for
specific purposes (LSP). However there are degrees of specificity, and he suggests the concept a "continuum of
specificity" (p 14) to explain the variety of LSP tests.
This brings us to a refreshingly concise summary of Douglas's (2000) ideas, together with the objections which
have been raised to them. Douglas's main concept is authenticity. An LSP test is "one in which . . . test tasks
and content are authentically representative of tasks in the target situation" (quoted on p 3). The two underlying
ideas are firstly that "language performance varies with the context of that performance" (p 3), and also that they
are distinguished from general-purpose tests in their language. Douglas says LSP tests are situationally authentic
because the language comes from a specific "language domain". They are also interactionally authentic, because the
tasks replicate target tasks.
O'Sullivan then attacks two main problems with LSP – and therefore also business English – tests: Can LSP
tests be clearly distinguished from general English tests? Can LSP tests be made both situationally and
interactionally authentic? His answer to the first question is that we know from studies of the role of
background knowledge in tests that specific areas are clearly definable, and that on a highly specific
test, subject experts outperform laymen.
As for authenticity, situational authenticity is possible, especially in highly specific tests such as for air
traffic controllers. Interactional authenticity is still a rather unsatisfactory concept. The test should result
in a interaction between the task and a linguistic skill, but "insufficient work has been done to link context-based
validity elements to theory-based processing" (p 7) This section concludes with a discussion of how to assess
performance by various rating scales.
The final part of the lengthy first chapter is a review of major business language tests, including those for
English, French, Italian and Japanese. A nine-point scheme is used to classify and evaluate the tests, and the
discussion is helped by a liberal use of charts and sample exam questions.
The most familiar of these tests to readers of this journal will probably be TOEIC®, which is
described as "problematic" for many reasons. It is too general to be classified as a business English test. There
is a contradiction between what the writers claim that the test shows – "TOIEC test scores indicate
how well people can communicate in English with others in the global workplace." (quoted on p 18) and what it really
allows us to say about the candidate. This is because "There is clearly some confusion as to the underlying construct
of the TOIEC" (p 18). Another problem is that the test is based on a theory of separate skills.
O'Sullivan criticizes the writers for "the failure . . . to respond to changes in theoretical perspectives of
language competence" (p 19) The skills coverage is inadequate because language knowledge rather then language
use is being tested. The score reporting is also unsatisfactory since the score is related to the population who
took that administration of the test, rather than to a criterion level.
[ p. 18 ]
The remainder of the text deals with the BEC test. Firstly, we read a history of the revision process, an
intriguing story for most of us who have never been involved in producing a large-scale test. The UCLES test
development method (described in Saville, 2002) is shown in practical terms. We are introduced to the construct
of business English and how it is arrived at practically by considering data on the test takers. The process of
assessing and using three types of validity are dealt with: theory-based validity, content-based validity and
scoring validity (reliability). Then the four qualities of the Cambridge ESOL tests – validity, reliability,
impact and practicability – are explained. Finally the practical issues of administration and grading are
addressed. The last few pages outline the revision process. This process was based on "an extensive consultative
exercise", (p 117) and examination qualities were identified using the VRIP system: "validity, reliability,
impact and practicality." (p 117)
Major changes to the suite is the title of the third, fairly brief, chapter. This deals with general changes such
as the way results are reported, the weighting of the components, both within each section of the test and also
within each BEC level, and the changes to the speaking content.
Following these high-level changes, the details and rationale for revisions of sections of the BEC test are analyzed
in Chapter 4. In short, "the most significant changes have come in response to criticism of the Speaking paper,
while the other papers . . . were left relatively untouched." (p 171).
The final chapter returns to an overview perspective. A welcome summary of
the book's essential argument is presented. This leads us to a discussion of specificity, authenticity and validity.
These theoretical discussions are balanced and clarified by a section on practical implications: reliability issues
and the use of computers. The author concludes with a thought-provoking agenda for LSP testing. For example, with
regard to testing of speaking, it is a good idea to include many different tasks. He suggests, "research is required
into the practicality of awarding scores of individual testing operational conditions." And he recommends, "It is
necessary to investigate the impact on performance of factors such as interlocutor variables (e.g. sex and age)
or candidates' perception of the interlocutor/audience . . ." (p 192)
The four appendices, running to nearly 200 pages, give illustrative sample papers of the pre-and post revision BEC
test in addition to other relevant tests: BULATS, CEIBT, and CEFL.
There is some reference to Japan, where the author worked for a number of years: the JETRO test of Business
Japanese is described. Moreover, eight pages are devoted to the TOEIC test. However, the language
testing world is very much an international community, and this is the overriding impression of this text.
The style is generally fluid, knowledgeable yet lucid, using the first person pronoun. In addition, the diagrams
are very helpful aids to exegesis of quite complex ideas. However, occasionally this reviewer stumbled in reading
and had to backtrack. This is due in part to the author's conversational writing style, whereby at the end of a
sentence an extra thought is added after a dash. This literary mannerism bloats many sentences, as for example
two successive sentences in the third paragraph on p 15. Moreover, perhaps a more energetically edited text would
have avoided double negatives, such as "a not incompatible expansion" on p 14. A final minor point there is at least
one run-on sentence, on p 192, the third complete sentence.
This text is recommended both to both specialists and generalists. Those who are involved in test production will
be enlightened by the theoretical discussion of LSP tests, and the rigorous analyses deployed in the case study
section of Chapters 2 - 4. Generalist teachers preparing students for the large-scale LSP examinations such as
the TOEFL® or TOEIC, will find this work both accessible and balanced. For native
English speaker students taking a testing module, it provides an up-to-date alternative to Douglas when considering
LSP testing.
- Reviewed by Daniel Dunkley
Aichi Gakuin University
Author IndexTitle IndexDate Index
Background
Links
Network
Join
[ p. 19 ]