This volume is an impressive demonstration of what is suggested by language testers' and teachers' discussions of test validity but is seldom undertaken. Hasselgreen's book represents a monumental and painstaking study that began with the pilot of the speaking portion of the evaluation of English as a school subject (EVA) in the spring of 1995 and, we can assume, has continued even after the publication of this book in 2004.
The book is divided into two sections: The first focuses on the identification of possible sources of test invalidity using a Messickian framework, and the second focuses on the role of “smallwords” and their contribution to fluency at three levels of proficiency (native-speaking high school students from the United Kingdom, high-proficiency Norwegian students, and lower proficiency Norwegian students). By concentrating so thoroughly on (a) the a priori application of a validity framework in the preliminary stages of test design and (b) the use of corpus-based analyses of the language elicited by the test in initial administrations, Hasselgreen's work represents a departure from what is often offered as evidence in support of test validity (e.g., applications of item response theory, examinations of rater behavior).
Hasselgreen comments in the introduction that the test is “fairly unremarkable” and that it is representative of “the type that anyone conversant with generally accepted principles and practice concerning the testing of spoken interaction, evolving around the turn of the millennium, might have produced, given resources and a relatively free hand” (p. 1). The EVA is a 20-min communicative test that is conducted in pairs and focuses on spoken interaction. It might be representative, as Hasselgreen seems to assume, of classroom-based assessment activities in English as a second language and English as a foreign language classrooms in Europe, in America, and throughout the world; however, what is remarkable is the availability of resources for test development, a relatively free hand in the development process, and the use of a pair-based interactive assessment at the national level. Hasselgreen comments:
The primary purpose of the testing was defined as providing teachers and students with detailed information on the strengths and weaknesses in students' communicative language ability (CLA) in English, so that learning activities could be better adapted to student' particular needs. The methods used were to be innovative and the process of taking part in the testing was to enhance the learning situation and the level of competence in assessment itself (p. 3).
This style of test development and national assessment—one in which the test and instruction are integrated and complementary—is remarkably different from the standard practice in the United States, where increasingly high-stakes national tests are authored by testing or publishing companies.
Thus, the Norwegian EVA represents an interesting compromise between large-scale and local testing systems designed to meet specific, program-related goals. In its deliberately non-high-stakes position in the testing-purpose continuum, there is clearly greater tolerance for problematic technical standards, but perhaps it is the shift away from the technical aspects of the test to its instructional value and the data elicited that allow Hasselgreen to address language learning, teaching, and acquisition.
The test becomes a springboard for the subsequent, primarily ad hoc, descriptive discourse-based analyses of examinees' fluency and smallword use at three levels of proficiency. Hasselgreen examines notions and discussions of fluency and links those discussions to examinations of what she identifies as smallwords (e.g., I think, you know, sort of, right, well, ok). These difficult words have been referred to elsewhere as collocations, lexical bundles, discourse markers, formulaic expressions, lexical fillers, and hedges—to name only a few of the more common labels in use. A criticism of Hasselgreen's smallword representation of the class is the lack of clear-cut, transparent categories for inclusion or exclusion. The definition she attempts, “small words and phrases, occurring with high frequency in the spoken language, that help to keep speech flowing, yet do not contribute essentially to the message itself” (p. 162), invites befuddling complications with even the slightest push. Hasselgreen acknowledges the difficulty, and it should be stated that it is a problem that she cannot reasonably be expected to resolve. What Hasselgreen does examine and develop in a way that no one has done before is the link between gains in fluency and gains in facility in smallword use. This is a book that should appeal to a broad audience. It is well worth reading and is highly recommended.
