Language for Specific Purpose (LSP) performance assessment in Asian call centres: strong and weak definitions

Abstract

LSP performance assessment has become a special focus for language testers in recent years where experts have debated how testing tools and processes can be strengthened to more accurately and more validly assess professional communication at work. Suggestions to achieve this include ethnographic studies of the target language situation; authentic discourse analyses of the relevant texts; and subject matter experts (SMEs) being invited as informants when defining ‘successful’ communication at work. These are proposed as key steps in building LSP performance assessment validity (see for example Jacoby & McNamara: English for Specific Purposes 18(3):203–241, 1999). Language testing researchers have also suggested there may be some merit in distinguishing LSP performance as ‘weak’ and ‘strong’ (see also McNamara: Measuring Second Language Performance, 1996; Douglas: Language Testing 18(2):171–185, 2001) although clear lines of distinction between these have not yet been made in LSP assessment studies to date. In this article I propose a distinction between a ‘weak’ and ‘strong’ version of LSP performance assessment based on the empirical data collected over the last ten years of developing and embedding the Business Performance Language Assessment Scales (BUPLAS) into Asian call centres. A distinction that has been unclear to date in this area of research is whether a workplace LSP spoken assessment for work (meaning those assessment tasks that gauge employment entry levels as ‘predictive’ of work success) is the same as a workplace LSP spoken assessments at work (meaning those authentic assessments that gauge quality levels on-the-job as ‘observed’ success at work). For the Asian call centre industry this distinction is important because the purpose of recruitment (exclusion) is very different from the purpose of quality assurance (appraisal and coaching feedback). I argue that recruitment assessment with its attendant characteristics constitutes a ‘weak’ version of LSP performance assessment whilst its quality assurance counterpart constitutes a ‘strong’ version.