Assessment Literacy for Language Educators part 5: The washback effect
In part 5 of this series focusing on Language Assessment Literacy, we will look at the ways in which language assessment can be harnessed as a tool for language development, and the ways in which test design can influence the quality of learning leading up to the test event. The relationship between testing and prior learning can be seen by investigating the washback effect, which can run through all aspects of assessment design for the benefit of the test takers.
The traditional testing paradigm
As everyone remembers from their own education, tests and exams can be nerve-racking experiences for students. Traditionally, learners have very little control over how they approach a test, and the construct is fixed, proscriptive and dictates what knowledge and skills need to be drawn on for success. Students may do a couple of mock exams or practice tests shortly before the real thing, and that is seen to suffice for them to understand the nature of the test, the types of activity they will be asked to perform and the thought processes that underlie the different question types. But is this enough? As an important part of development, shouldn’t assessments work in line with the learning and upgrades in the subject being taught, as a learning tool? This would ease the pressure on the test-takers, inform them more deeply about why they perform the activity that they do on the test, and may even lead to greater learning through the periods of study leading up to the assessment. In order to work towards this kind of more positive assessment approach, it is important to understand the different ways in which testing can affect learning: the washback effect
What is washback?
Put simply, washback is any effect which the exam situation at the end of a period of study has on the nature of the study itself. This includes student and teacher assumptions, feelings, thought processes, preparatory activity, types of teaching and learning which go on because of the nature of the upcoming test, which ‘wash back’ from the test event itself over the teaching and learning activity which is performed during the course of study. Washback effects can be positive or (most commonly) negative, with the stresses of revision, time pressure to prepare towards the all-important date, and anxiety about the result affecting students and teachers alike. In language learning, courses of study are often designed to focus only on items which appear on the test, leaving significant aspects of language use untaught; if the test includes reading, writing and listening activity, but has no face-to-face speaking component, it is logical to assume that oral communication skills will get deprioritised by teachers and learners, despite the importance of this essential language area. This is exactly what happens in most language testing environments in the world, meaning that areas of language development may suffer as a direct result of the assessment construct. This is an example of a major negative washback effect which affects millions of learners’ language every year.
Negative washback effects
Aside from negative washback from the broad skills which are assessed in an exam, other more specific types of negative washback are often seen in traditional language assessment. Very proscriptive exam types may include only items which can be empirically measured as ‘correct’, such as multiple choice items, exact expected answers and one-word responses. Including only these types of item in language assessment can be dangerous in that it reduces the scope for free expression, and preparatory classes may well take on that tone, with teachers seeking out only ‘correct’ responses from students, and labelling different, perhaps even more creative or communicatively effective, responses as ‘wrong’. This can affect language performance significantly, and is one reason why language users from certain cultures are much more reticent to speak out than others. This style of test can also prioritise accuracy over fluency, meaning that it becomes difficult for students to speak, listen, read or write at length as unknown words or errors in production can become unsurmountable barriers to language use.
Another negative effect which can wash back from assessment design is a focus on language as a knowledge-based subject. If an exam asks students to show how many words they know directly, or treats language as an information set that must be demonstrated through multiple choice questions and short answers only, knowledge about the language studied may become more important than the message being conveyed by the test-taker. In restricted preparation for this kind of exam, it may become difficult for students to cope with ideas that they are not already familiar with, or to deal with words that they do not already know the meaning of, both of which are incredibly important skills in a learner of a foreign language. It is simply unrealistic for a learner to only meet words and structures that they have already studied to familiarity.
So how can we design tests to have positive effects on learning?
Positive washback will occur if the construct of an exam mirrors the range of skills, language knowledge and language systems which, during preparation for the exam, will promote a holistic set of communicative competencies, and include not just grammar, vocabulary and four-skills work, but also require test-takers to achieve free-expression tasks using the language that they have at their disposal. The freer the task, the more information can be gained about how the learner processes their language, and the language choices that they make given the chance in a less proscriptive environment, just as we do in the real world. Of course, there must be strict parameters to define the quality of language being tested, but these can be stated so as to take into account the full range of linguistic, sociocognitive and strategic competencies which any language user employs in their own language every day. Considering language performance in this broader way enables a truer picture of test-takers’ language and can lead to much more positive, empowering washback effects in preparatory activity.
Part 6 of this series will investigate how positive washback can be achieved through assessment design and supported teaching, and the ways in which learnt competences can be aligned with the test construct to enable greater all-round performance in the language being tested.
Tom Garside is the founder of Language Point Teacher Education Ltd., an education development organisation which leads professional development and teacher training activity with the aim of empowering teachers in the contexts where they work and increasing sustainability to the development activity that they participate in. He also delivers Trinity CertTESOL courses to English language educators around the world.