Language testing and technology

In this chapter we discuss how the purpose of a language test is to take a snapshot of a person’s language ability and use that to infer what the person is likely to be able to do with the language in a wider range of real-life situations. The snapshots are limited in time and number of shots that can be taken, so they can only capture a small part of what the person can do in the language. Test developers, therefore, strive to design and administer test tasks which represent the types of activities that test-takers are likely to encounter in real life – as much as possible within the practical constraints. Producing a language test typically involves several stages, such as specification, item and task writing, administration, scoring, and score report production Today, technology is applied at all stages. To closely review and discuss the widespread use of technology in test development and validation in this chapter, we draw on Weir’s (2005, Language Testing and Validation, Palgrave) socio-cognitive framework. Although there are multiple frameworks for evaluating test validity, the usability of the socio-cognitive framework has been testified to by various testing organisations worldwide. It should also be noted that this chapter mainly refers to large-scale English tests, such as IELTS and the Duolingo English Test, due to the availability of published information, but implications for language classrooms are also discussed where relevant.