An external test or public examination, particularly in a second/foreign language such as IELTS, TOEFL, HSC examination in English, etc might have considerable impact on the stakeholders including teachers, learners, parents, administrators, the institution, the educational system and society as a whole. The effect a test has on the different components of the educational process of a second/foreign language is termed ‘washback, either beneficial or harmful at both the micro and macro level. The current paper, firstly, purports to be an appraisal of the concept of ‘washback’, secondly, examines its nature and functions as revealed by different proponents and researchers, and, finally, ascertains its role in second/foreign language education.
Key words: second/foreign language tests, washback, origin of washback, positive washback, negative washback, washback hypotheses, stakeholders, micro and macro level, minimizing negative washback and maximizing positive washback
A public or external examination such as IELTS, TOEFL, HSC examination in English and so forth is generally considered as an external test since it is planned, designed, arranged, administered and scored by external agencies or forces such as education boards, state departments and so on with a view to evaluating learning products or outcome with a decisive consequence for or influence on examinees or testees. Such tests are often exploited as tools to select students as well as a means to control an educational system of an institution, especially when the educational system is driven by tests or examinations (Cheng and Falvey, 2000; Cheng, 2009; Herman, 1992; Smith et al, 1990), and commonly assumed to have an impact on both teaching and learning a second/foreign language. Given that external tests or public examinations have exerted an influence on teachers and students with an associated impact on what occurs in the classroom, such a phenomenon is represented as ‘washback’ or ‘backwash’. In other words, an external test or public examination influences the attitudes, behaviours and motivation of teachers, learners, and parents; and as it often comes at the end of a course or programme, this influence is seen working in a backward direction, and hence emerges the term ‘washback’ in second/foreign language education.
The origin of public examinations is to be found in the school entrance and civil service examinations of China, which go back at least to the period of the Sui emperors (589-618), and which achieved their most complex form towards the end of the Ch'ing dynasty (1644-1911) (Miyazaki, 1976). Inspired by the Chinese systems, examinations in written format began to appear in European schools in the 16th century, though it was not until some two hundred years later that public examinations of the type found in China were instituted in Europe for selection to universities, the civil service, and the professions. Public examinations are then a major feature of the educational systems of most European countries, which, in turn, passed them on to their former colonies in Africa, Asia, and the Caribbean, where they still flourish (Kellaghan, 1992). The United States, with some exceptions (e.g. the Regents' Examinations in New York), has so far not adopted a public examination system. However, during the 1980s and 1990s, a number of proposals contained in reform reports, policy statements, and legislation advocated a national system or systems of examinations for the country (Madaus & Kellaghan, 1991).
Although there is considerable variation in the form and administration of examinations from country to country (Madaus & Kellaghan, 1991; Noah & Eckstein, 1992), they generally share a number of characteristics (Kellaghan, 1993). Firstly, the examinations are controlled to varying degrees at national or regional level (and sometimes also administered) by an agency or agencies outside the institution, usually a state department of education, an examinations council closely related to the state department, or regional examining boards. Secondly, the examinations are geared to syllabi which are usually defined by an agency outside the institution, sometimes the same agency as administers the examinations. Thirdly, examinations are usually provided in the traditional areas of the curriculum such as mathematics, science, language. Fourthly, examinations are often formal terminal procedures taken on fixed days under controlled conditions by all the candidates of the examination in a country or region at the end of a course of study. Fifthly, examinations are largely written, very often using the essay format, but sometimes making use of multiple-choice items, either in conjunction with other formats or on their own. There may also be provision for oral and practical assessments. Finally, as a result of performance in the examination, the student is awarded grades or marks in each subject examined.
Public examinations are normally intended to serve a number of functions. The most obvious is to assess the competence of students’ learning relative to some agreed standards. The results are then frequently used to discriminate among students with regard to their preferred futures: further education, admission to professional preparation, or employment. While certification is important, particularly for students who are leaving the educational system, there is often a danger of losing sight of this function because of the strong emphasis on selection. Examination results are also often used, formally or informally, to provide evidence of the effectiveness of institutions, and institutions and teachers may be held accountable for their students' achievements as reflected in examination performance. This use becomes more obvious when results for individual institutions are published. Washback evidently indicates an intended or unintended direction and function of curriculum change on aspects of teaching and learning by means of a change of public examinations.
The present paper is, therefore, intended to (a) appraise the term ‘washback’ of tests; (b) examine the nature and functions of washback as disclosed by different exponents and researchers; and (c) ascertain the role of washback in second/foreign language education.
Background and Origin
‘Washback’ which was rarely found in the dictionaries published before the 1990s is a term now commonly used in applied linguistics. Washback on learners was a topic seldom discussed in the 1990s, and has received more attention from researchers since the start of the 21st century. The argument that the quickest way of changing students’ learning is to change the assessment system has been widely accepted.
Test impact on society and individuals has been the subject of considerable research interest in the field of language testing during the last two decades (Alderson and Wall 1993, 1996, Bailey 1996). The social consequences of test results are regarded as a central aspect of construct validity according to Messick’s (1989) definition, and the focus on test impact is claimed to distinguish modern language testing in the communicative paradigm from language testing before the 1970s (Bailey 1996).
The first washback study covering test impact was conducted by Kellaghan et al. (1982), then Wesdorp (1982) and Hughes (1988). It should be pointed out that the former was a general education study and not specific to language education. In their ensuing discussion, it is clear that evidence of either beneficial or harmful was often tenuous remaining unproven or, at best, inconclusive. For example, the study conducted by Kellaghan et al. (1982) looked at the impact of introducing standardized tests in Irish Schools as a case in point. It is often suggested that because test information is important in attempting to hold schools accountable, the influence of tests on what is taught is potentially great (Gipps, 1994). Nearly 25 years ago, Alderson (1986) identified washback as a distinct and emerging area within the field of language testing. At around the same time, Davis (1990) was asking whether tests should necessarily follow the curriculum, and suggested that perhaps tests ought to lead and influence the curriculum.
Afterwards, ‘washback’ on learners was a topic discussed with interest during 1990s, and has received more attention from the researchers since the 21st century. The Sri Lankan impact study conducted by Wall & Alderson (1993) is often cited as a landmark study in the investigation of washback. Washback or backwash, as it is sometimes called, is now a term that is commonly used in the assessment and applied linguistics literature (see Alderson & Wall, 1993; Biggs, 1995).
When the relationship between testing and society is explored in second/foreign education, the focus is almost exclusively on the impact of testing on society. It does, however, seem a reasonable assumption that there is a two-way relationship between testing and society: not only do language tests affect society, language tests are also affected by society. The kind of society of which tests are a part affects test development, testing policy, the use of tests as well as public opinion about tests. At present, washback studies are being conducted throughout the world as an emerging discipline. Now, a slogan is being echoed “testing for learning, not learning for testing”.
A test or examination, an unavoidably pervasive part of a second/foreign language education system, refers to a formal situation created by the tester to make the testee respond to a stimulus from which intended information could be elicited so as to determine the testee’s ability or knowledge, and/or to grade him/her according to an established standard. Therefore, it is believed to be a powerful determiner of what happens in the language classroom, and is claimed that it affects teaching and learning activities both directly and indirectly. It has long been asserted in a wide range of contexts that a test exerts a powerful influence on the language learner who is preparing to take the test and on the teacher who tries to help him/her prepare. It is generally accepted that public examinations or external tests influence the attitudes, behaviour, and motivation of teachers, learners and parents (Pearson, 1988). The influence that a test has on teaching and learning a second/foreign language is commonly termed as ‘washback’ in applied linguistics (see Cheng and Curtis, 2004; Gates 1995; Alderson and Wall, 1993).
The concept of washback sometimes referred to as backwash (Biggs, 1995, 1996 cited in Cheng, 2000) can generally be considered as the impact of an examination or test on teaching and learning (Cheng, 2003, 2008). Similarly, Brown (2000, p. 298) defines washback as “the connection between testing and learning”. Alderson and Wall (1993, p. 118), however, restrict the use of the term ‘washback’ to “classroom behaviors of teachers and learners rather than the nature of printed and other pedagogic material”. They also regard washback to be what teachers and learners do that “they would not necessarily otherwise do” (p. 117). Messick (1996, p. 16) states that in order to be considered washback, good or bad teaching has to be “evidently linked to the introduction and use of the test”. Moreover, Wall (1997) makes a clear distinction between washback and test impact. The latter would refer to the effect of a test on “individuals, policies or practices, within the classroom, the school, the educational system or society as a whole” (cited in Cheng and Curtis, 2004, p.4). Other researchers (Andrews, Fullilove & Wong, 2002) do not make this distinction, and maintain that both narrow and wider effects of tests can be included under the term ‘washback’. It is common to claim the existence of washback and to declare that tests can be powerful determiners, both positively and negatively, of what happens in classrooms (Wall & Alderson, 1993). Swain (1985, p. 43) succinctly states the prevailing opinion: "It has frequently been noted that teachers will teach to a test: that is, if they know the content of a test and/or the format of a test, they will teach their students accordingly".
Tests are often perceived as exerting a conservative force which impedes progress. Andrews and Fullilove (1994, p. 57) point out, "Not only have many tests failed to change, but they have continued to exert a powerful negative washback effect on teaching”. These authors also note that "educationalists often decry the 'negative' washback effects of examinations and regard washback as an impediment to educational reform or 'progressive' innovation in schools" (ibid., p. 59-60). Likewise, Heyneman (1987, p. 260) has commented, "It's true that teachers teach to an examination. National officials have three choices with regard to this 'backwash effect': they can fight it, ignore it, or use it".
Hughes (1989, p.1) states “the effect of testing on teaching and learning is known as backwash" (and this term, as he uses it, is synonymous with washback). Further, Pierce (1992) states the washback effect sometimes referred to as the systemic validity of a test. Alderson & Wall (1993, p.1) suggest washback compels “teachers and learners to do things they would not necessarily otherwise do because of the test”. Shohamy (1993, p. 4) summarizes four key definitions that are useful in understanding the washback concept: (a) Washback effect refers to the impact that tests have on teaching and learning; (b) Measurement driven instruction refers to the notion that tests should drive learning; (c) Curriculum alignment focuses on the connection between testing and the teaching syllabus; and (d) Systemic validity implies the integration of tests into the educational system and the need to demonstrate that the introduction of a new test can improve learning.
Cohen (1994, p. 41) describes washback in terms of "how assessment instruments affect educational practices and beliefs". Andrews (1994, p.45) sees washback as "an influence on teachers, learners, and parents, with an associated impact on what happens in classrooms". Gates (1995, p.101) defines washback simply as “the influence of testing on teaching and learning”. Bachman and Palmer (1996) have discussed washback as a subset of a test's impact on society, educational systems, and individuals. According to Messick (1996), washback is the extent to which the test influences language teachers and learners to do things that they would not necessarily otherwise do. Bailey (1996, p.5) states, “washback is the influence of testing on teaching and learning”. Shohamy, et al. (1996, p.6) maintain that washback is delineated as “the connections between testing and learning”. Likewise, Cheng (2005, p.8) contends that washback indicates “an intended or unintended (accidental) direction and function of curriculum change on aspects of teaching and learning by means of a change of public examinations”.
Thus, definitions of washback are nearly as numerous as the people who write about it and who investigate it. These definitions range from simple and straight-forward to very complex. Some take a narrow focus on teachers and learners in classroom settings, while others include reference to tests' influences on educational systems and even on society in general. Some descriptions emphasize intentionality while others refer to the apparently haphazard and often unpredictable nature of washback. Notwithstanding, it is certain that washback exists, and affects the diverse facets of the whole educational process of the language programme encompassing the teacher, the learner, the institution, the administration, and the social setting as a whole.