ABSTRACT
In language for specific purposes (LSP) testing, the task of the test designer is to mirror as accurately as possible the language, tasks, and contexts of the target language situation in order to accurately predict workplace language performance. In such performance-based testing, best practices demand that the representativeness of the test tasks and the language required be based on a needs analysis of the target language use situation (McNamara, 1996; Douglas, 2000). For work-related LSP tests, the target language use situation is most often framed within a
communicative competence model (Hymes, 1972; Canale and Swain, 1980). In this approach, appropriate language testing requires test takers to engage in extended acts of communication in which they assume social roles and perform authentic tasks common in appropriate real-world settings. In order to determine what roles and tasks must be included in the LSP test, a careful analysis of the work domain is essential. The needs analysis requires a focus on interaction in the work setting to determine what kind of exchanges typically occur, who the participants are, what the physical setting of the interaction is, what the purpose of the interaction is and what communicative functions and linguistic forms are used. However, language testers should also be cognizant of the social and political ramifications of test design and consequent decisions. What is the larger context in which the testing is framed? Who is to be tested and why? What societal pressures have established the need for the test? What cultural norms and values are realized in the formulation of the standards? In high stakes situations, where testing decisions have major consequences for the test takers and for the general public, understanding the sociopolitical underpinnings of a test situation is essential to ethical and fair test design (McNamara and Roever, 2006). This chapter traces assessment design practices within a socio-political context for language tests
that are employed in work-related contexts where the language competency has direct effects on a large segment of society. We will outline the procedures for determining the target language to be assessed: gathering information from stakeholders, consulting work-related documents and manuals, examining discourse-based research, observing workplace task and language use, corpus generation and analysis, and piloting sample items with various stakeholders. Throughout our discussion we will consider the ways in which social and political concerns affect testing decisions. The domain in
which we will illustrate these practices is Aviation English. Aviation English is of particular relevance for our purposes because of the socio-political context in which the language proficiency standards emerge and the distinctiveness of the language required in the domain. The use of English in aviation is governed by the International Civil Aviation Organization
(ICAO), a specialized agency of the United Nations, which creates standards and recommendations governing civil aviation “by and for” its 190 contracting member states (ICAO, n.d.). The inherently political nature of a test situation in which 190 member nations with varying interests and resources must debate and agree on a common policy that balances workplace needs with public safety makes the testing of Aviation English an excellent example of the social uses of language tests.
