Dr Evelina Galaczi of Cambridge Assessment English explains why language assessment should reflect practical applications. How can exams cover the communication skills and mental processes required in real life?
Research on the impact of tests on learning indicates that tests exert a strong influence on classrooms (test washback) and education policy in general (test impact). Good tests lead to good teaching, which leads to good learning. For example, a test which includes a speaking component will likely lead to classrooms which focus on developing communicative skills. As a teacher preparing students for a Cambridge English exam in Spain said, ‘We have more communication activities as a direct result of the exam’ (Aston 2012:39).
Research from Cambridge Assessment English impact studies has shown that preparing for communicative English tests has several positive benefits:
- Learners use much more English in schools
- Learners’ language proficiency improves
- Teacher motivation increases
- Teaching practice improves
Authenticity is a key element of assessment. A test is authentic if it resembles real-life situations/tasks and triggers mental processes similar to the same task in a real-life setting.
Situational authenticity refers to how closely a task in a test resembles a real-life task. For example, asking the test-takers to write an essay in an Academic English test would likely have situational authenticity, while asking them to write a story would not, since writing an academic essay is similar to the tasks those students would encounter outside of the classroom or test.
In addition, it is important that a test task has interactional authenticity, so that the mental processes required by the task are similar to the ones in the corresponding task in real life. Tasks that have interactional authenticity may not necessarily replicate a real-world context, but they elicit a cognitively authentic linguistic experience since they create a context in which realistic uses of language can occur. For example, test-takers might be asked to look at two photographs and compare them. While it’s unlikely that this situation would arise in real life, the ability to perform this kind of monologue taps into features of real-life speaking, such as organising information in a logical way, using discourse markers to signal contrast and comparison, and spontaneously producing a range of vocabulary grammar structures.
A test needs to include a range of tasks, which replicate the features of non-test tasks as much as possible (situational authenticity) and allow opportunities for different mental processes to be activated (interactional authenticity). This allows closer approximation to real-life skills.
When a mother tells a story or when a professor delivers a lecture, they are transferring information in a monologue. When friends are chatting or workers are having a meeting, they are exchanging information. As far as possible tests need to include tasks which elicit different features of speaking and include monologue speech and interaction. For example, the Cambridge English First exam includes four different types of tasks: 1-to-1 interaction between the examiner and each test-taker; a monologue task for each test-taker; a paired interaction task between the test takers; and a group interaction task between the examiner and test-takers.
Real-life speaking depends not just on grammar and vocabulary, but on using them to talk in a way that is intelligible, makes logical sense, flows well and works as part of a conversation. These requirements should be reflected in the assessment criteria in order to ensure that all aspects of speaking which matter in real-life communication are given appropriate weight, and categories such as grammar and vocabulary, although fundamental to speaking, are not weighted excessively.
Real-life listening takes place in different situations and places demands on different levels of mental processing. For example, when we listen to a person talking on the radio or to people discussing something in a meeting, we are not participants, just listeners (receptive listening). In conversations, however, we have to be both listeners and speakers (interactive listening). Good tests involve both these types of listening.
Real-life listening also involves visual cues, which provide support as listeners decode the message. This is not fully possible in listening tests, but a brief rubric at the beginning of each section can compensate for the absence of visual support and provides some authenticity. Listening also involves a range of voices, accents and speech rate, and that should be reflected in tests.
There are different types of reading – we might read a book for pleasure, a newspaper for information and a letter to get an update from a friend. We also read at different levels of attention/mental processing: sometimes we read carefully, e.g. a book chapter on a topic we are interested in, and sometimes we read expeditiously, e.g. skimming a movie review to get the gist.
When we read we sometimes focus only on the meaning of individual words (‘local reading’), but typically we read beyond the meaning of individual words and sentences to understand the general meaning of entire paragraphs (‘global reading’). This involves different types of mental processes, and good reading tests include a range of tasks tapping into both expeditious and careful reading at the local and global level.
Real-life writing involves the ability to produce different kinds of writing. This involves simple writing such as completing a form or writing a brief e-mail to writing reports, letters, stories, articles, etc. Different writing tasks (genres) are needed in good writing tests; at lower proficiency levels test-takers could write an e-mail and a note to a friend, at higher levels they might have to write an essay for a teacher and either a book review or an article. This range of tasks adds to the authenticity of writing exams.
The assessment categories for writing should also reflect the important features of real-life writing. For example, they should focus not just on the range and accuracy of grammar and vocabulary, but also on how well the task been completed, how it’s organised, and how appropriate the style and register is for the task.
Dr Evelina Galaczi
Head of Research Strategy, Cambridge Assessment English
Before you go... Did you find this article on Twitter, LinkedIn or Facebook? Remember to go back and share it with your friends and colleagues!