Blurb:

Chapter 2 discusses the use of Weir’s (2005) socio-cognitive framework to systematically guide speaking test development. It highlights the importance of a priori validation as an essential starting point of test development projects. It shows how a series of studies on context and cognitive validity can inform the development of test specifications. It also focuses on the development and validation of rating scales based on empirical evidence from test-takers’ output language, as well as describes different rating systems and their pros and cons. This chapter exemplifies some large-scale test development projects, such as Trinity’s Integrated Skills in English (ISE) examinations, Eiken’s Test of English for Academic purposes (TEAP), and the British Council’s Aptis test. Chapter 2 also demonstrates how these tests, that addressed context and cognitive validity at the early test development phases rather than relying solely on a posteriori validation, can respond to the concept of ‘impact by design’, and contribute to building a bridge between large-scale high-stakes tests and classroom assessment, as discussed in Chapter 1.