Backwash Effect

The backwash effect: fromtesting to teaching

Luke Prodromou

Teach: If you teach someone something you give them instructions sothey know about it or how to do it; you make them think, feel or actin a new or different way; you explain or show students how to dosomething. (Collins' COBUILD Dictionary)

Test: To find out how much someone knows by asking them questions.(Longman's Active Study Dictionary).

'Teach' and 'test' are quite close together in a dictionary, but in testing wedo different things from the things we do when we teach. This articleassesses the concept of 'backwash' in language teaching, looks at theconsequences of testing on teaching in a broad educational context, andsuggests that 'negative backwash' makes good language teaching moredifficult. The two processes of testing and teaching are considered to benecessary but distinct. A system is described for distinguishing betweenthem which is then applied to developing classroom activities forexamination preparation classes, to help teachers move from testing toteaching procedures.

What is the The backwash effect can be defined as the direct or indirect effect ofbackwash effect? examinations on teaching methods. According to the effect of

examinations on what we do in the classroom we may refer to 'positive'and 'negative' backwash (Heaton 1990: 170, Hughes 1989: 1). Althoughit is an important factor in classrooms wherever examinations play adominant role in the educational process, it has not been fully explored. Itis not mentioned in the index to such standard ELT handbooks as Stern(1983), Howatt (1984), or Harmer (1991), and reference books such asRichards et al. (1985) and Seaton (1982) do not consider it worthy of anentry. Heaton (1990) and Hughes (1989) discuss, rather sketchily, what Irefer to as 'overt' backwash (see below), but do not explore the broadereducational implications of 'covert' backwash. The most thoroughtreatment of the concept of backwash is that of Alderson and Wall (1993),who suggest that 'washback' as they call it, is more complex than hashitherto been assumed. They make the valid point that there is no one-to-one relationship between tests, good or bad, and their effect on theclassroom. In their view, before a test has any impact on classroompractice it is mediated by factors such as the place of examinations inparticular societies, the teacher's competence, and the resources availablewithin the school system.

ELT Journal Volume 4911 January 1995 © Oxford University Press 1995 13

by guest on February 13, 2011

eltj.oxfordjournals.orgD

ownloaded from

http://eltj.oxfordjournals.org/

Theconsequences of

backwash

Whether the backwash effect is positive or negative, how it operates inparticular contexts—indeed, whether it exists at all—must be exploredempirically. Many of the assumptions about backwash are untested andsimplistic. Alderson and Wall (1993) point out that very little observationof the effect has been carried out, and that what evidence there is points tothe highly complex nature of the process.

Bearing these words of caution in mind, it might be useful to provide abrief background to the description of backwash put forward in thisarticle. I have been involved in examinations at several levels: as ateacher, trainer, examiner, and writer of tests and examination-relatedmaterials. The backwash effect described here is based on my observationof examination classes in the private and public sectors, over a period oftwenty years, in a society (Greece) where examinations play a verysignificant role.

Professional neglect of the backwash effect (what it is, how it operates,and its consequences) is one of the main reasons why new methods oftenfail to take root in language classes. Many teachers, trapped in anexamination preparation cycle, feel that communicative and humanisticmethodologies are luxuries they cannot afford. When the market calls onteachers and institutions to produce quantifiable results, it usually meansgood examination results. Sound teaching practices are often sacrificed inan anxious attempt to 'cover' the examination syllabus, and to keep aheadof the competition. In summary, 'negative backwash', as experienced bythe learner, means language learning in a stressful, textbook-boundenvironment.

The value oftesting

Uses and abusesof testing

It goes without saying that tests and examinations—at the right time, inthe right proportions—have a valuable contribution to make in assessinglearners' proficiency, progress, and achievement. As a device fordiagnosing learners' errors, and for defining the interlanguage ofindividuals and groups of learners, they are indispensable. Tests are alsothe simplest and most effective form of extrinsic motivation, of imposingdiscipline on the most unruly class, and of ensuring attention as well asregular attendance. Because they are closely bound up with classroomauthority, tests invariably lead to teacher-centred lessons, especiallywhere the teacher is inexperienced or insecure.

Abuse of testing occurs when tests invade essential teaching space, whenthey are not the final stage of a process of learning but become thebeginning, middle, and end of the whole process. Testing may be a shortcut to extrinsic motivation, but constant resort to it is an admission of theteacher's failure to make intrinsic motivation work. In the long run, it willdemotivate the learner.

Overt backwash The backwash effect can be overt or covert. In its overt forms, it usuallymeans doing a lot of past papers in class as preparation for an examination;it may involve replicating, from past papers or the textbook, the exercise-

14 Luke Prodromou



ownloaded from


types favoured in the particular examination students will be taking:multiple-choice, transformation, or gap-filling. The methodological routinethat results from the negative backwash effect in its overt forms is an all toofamiliar one: presentation of the text followed by questions similar to thosein the examination. This 'text + questions' formula is a crude mirror-imageof what happens in most conventional examinations.

Other hallmarks of the backwash effect include the use of fragments of(often inauthentic) language, a concentration on word- and sentence-levellinguistic features, and a focus on skills which in terms of administrationand marking are easier to test. This is why reading and writing tend to begiven much more emphasis in classrooms than speaking and listening.

This kind of overt backwash is usually negative, but there is no reasonwhy we should not have tests which adopt techniques more in line withcommunicative and, to some extent, humanistic teaching. Fortunately, itseems that most examination boards are aware of the problem, and aretaking steps to tip the balance in favour of positive backwash. It is possiblefor testing procedures to have a positive effect on classroom practice. Forexample, when one of the examination boards introduced a listening testbased on audio-cassette material (to replace the texts read aloud by anexamination supervisor), this had the effect of heightening awareness ofwhat authentic listening involves, and schools quickly began to preparestudents to cope with the new challenges.

Covert backwash The explicit consequences of the backwash effect are easily identifiable.The implicit consequences are more elusive, and more disturbing. Even ifexamination boards reduced the number of boring multiple-choiceexercises, the examination class would still be in conflict with theteacher's desire to teach communicatively and humanistically. This isbecause covert testing will always be with us. It is a deep-seated, oftenunconscious process, which reflects unexamined assumptions about awide range of pedagogic principles: how people learn, the relationshipbetween learner and teacher, the nature of teacher authority, theimportance of correction, the balance between form and content, the roleof classroom management, and so on.

Basically, covert testing amounts to teaching a textbook as if it were atestbook. Usually the teacher is not fully aware of this process: in his orher mind there is a clear dividing line between a lesson which involvesteaching and one which involves testing. I am using the latter term in aspecific sense which includes both overt and covert backwash effects.Some examples of covert testing will show what I mean. I have observedmany lessons where the teacher asks a question, receives a correct answerfrom a particular student, and then moves on to ask the next student thenext question. The objective of this routine is to find out what the studentsknow. This, and the lack of involvement of the rest of the class in thesequence, makes the activity more of an informal assessment than ateaching procedure. The absence of any lead-in or follow-up to the workdone on a text is entirely typical of testing procedures.

The backwash effect: testing and teaching 15



ownloaded from


Lead-ins and follow-ups have become standard teaching devices (see Peck1988: 201, where he refers to them as 'heads and tails'). The pedagogicrationale of a lead-in is to arouse interest and draw on the students'knowledge, thereby making learning 'easier'. By drawing a personalresponse from students, a follow-up will help fix or anchor the new input inthe learners' memories. A good teacher maximises the learners' chance ofsuccess by pre-teaching vocabulary, and doing pre-Jistening and pre-reading tasks to motivate learners, activate their past experience, and drawon their potential for more effective learning strategies. This approachcould not be more in contrast to the standard ritual in classroom tests andpublic examinations, where the teacher simply gives out the papers, andinstructs students to 'get on with it' in silence.

Penalizing error Testing values correct answers, and penalizes error. But in teaching weshould be as interested in the process by which students arrive at thewrong answer as we are in the correct answer itself. Holt (1964: 142-3)described how the process of 'only the right answer' ignores the stageindividual students have reached in their learning, and imposes on themmodels of performance based on the 'good' learners in the class. Intesting, the good learner is a yardstick by which all students are measured;in teaching, the student is his or her own yardstick. This is important:covert testing occurs whenever we do not give individuals their own spaceand time to answer questions; it is in subtle, invisible ways like this thatwe set up students to fail. Failure may be an inevitable feature of thediscrimination required in testing procedures, and the classroomhierarchies this leads to; in teaching, however, discrimination in thenegative sense has no place—for the good language teacher, success intests should be as routine as failure.

Asking questions In overt tests, the teacher or the examiners ask a lot of questions, butstudents taking public examinations, for instance, are expresslydiscouraged from doing so, unless there are exceptional circumstances. Incovert testing, too, the teacher asks a lot of questions, while the studentsare not given much opportunity to ask questions (of the teacher or eachother). A teaching procedure, on the other hand, allows students toexercise the power of asking questions; question-asking is accepted as anassertion of personality that can give a boost to self-confidence. It issymptomatic of the psychology of conventional testing that questions arediscouraged, and worrying to note how often teaching mimics the mono-interrogative mode of public examination, with parallel systems ofteacher authority and student submissiveness.

Denying learners' In covert testing less able learners are penalized by the collectivethinking time assumption that the objective of teacher questions is to elicit the right

answer in the shortest possible time. Thus, good students shout out theanswers, put their hands up first, fill in the pauses created by 'slower'learners searching for the right answer. Testing abhors pauses, which itsees as a vacuum rather than a necessary space in which students find theirown level.

16 Luke Prodromou



ownloaded from


The proxemics ofcovert testing

Denying learnerscommunication

Inflation of teacherauthority

Some learners need more thinking time than others, but conventionaltesting conditions impose a strict time limit on the production ofknowledge. It is not uncommon to hear of highly intelligent people whohave failed public examinations because of time constraints, or aninability to adapt their learning style to examination conditions.

In denying learners essential thinking time teachers often unconsciouslyrecreate these conditions; it is a great temptation to accept a correctanswer from the quicker students and move on to the next question. Notgiving students the time they need to prepare and process language, eitherin whole-class work or in pairs, creates anxiety, even panic, and thereforeerror—teacher-induced error. The strict time limits of formal tests canproduce error in the same way.

Covert testing routines are often accompanied and reinforced by theteacher's approach to classroom management. The use of space is oneimportant dimension of the management of groups. Teachers can teachbadly, not because of the methods or techniques they have adopted, butthrough mismanagement of space.

Many teachers tend to move closer to the student they have asked toanswer a question, and to fix their gaze on this student as they wait for theanswer they have in mind. This proximity of teacher and nominatedstudent tends to exclude the rest of the class; the teacher's body language,almost invites the non-participants, to 'switch off and talk amongstthemselves, which they often do, till the teacher turns to them in search ofthe next correct answer.

A powerful visual message is also conveyed by the way desks arearranged in the classroom. In testing, the desks are invariably arranged instraight lines with a space between them large enough to deter studentsfrom communicating with each other. Communication between studentsin a test is thus both implictly and explicitly forbidden.

In teaching, by contrast, we encourage sharing and communication byarranging desks in a semi-circular or group formation. These are familiardichotomies. Yet how many teachers go into a classroom for an ordinarylesson where the desks have already been laid out in linear fashion and leavethem exactly as they are, even though a horse-shoe or group arrangement ispossible? The learners are thus given an unspoken but powerful messageabout the teacher's methodological assumptions; what in teaching we wouldcall 'caring and sharing', in testing becomes 'cheating'.

The way we use space in class is as important as the texts we choose andthe methodology we adopt in presenting them. An arrangement of desks,appropriate in the context of objective assessment, when transferred toeveryday teaching, may obstruct the process of learning.

Testing, overt and covert is, as Fabian (1982: 24) has argued, apaternalistic, teacher-centred business: 'Examinations—like democraticinstitutions—do not thrive in isolation. When the consumer and the




ownloaded from


Characteristics oftesting and

teaching

Failure andsuccess

community at large surrender to academic technicians their right and dutyto be involved, they also surrender their right to check on the teachingstrategies that are the direct result.'

The premium placed on the 'right answer' in both overt and covert testinginevitably adds to the inflation of the teacher's authority, based on his or herrole as arbiter of correctness. One of my main arguments is that we need tomove away from this relationship towards a learner-centred approach totesting, and I will give examples of how this can begin to happen.

In Table 1,1 summarize what I feel are the most important characteristicsof the teaching and testing processes.

The qualities listed there under 'Teaching' are based on my ownobservation of teachers, native-speakers and non-native speakers, in thecontext of a number of teacher training courses and on a survey I havebeen conducting with students into what makes a good language teacher. Ihave also drawn on the work into effective teaching reported in Holt(1964), Peck (1988), Richards (1990), and Harmer (1991).

The features listed under 'Testing' are those we normally associate withthe backwash effect, with the addition of what I have referred to as covertsymptoms of backwash. While most public examinations and tests, oftenagainst the testers' wishes, encourage attitudes to learning summed upunder Testing, it should be said that a number of recent publicexaminations have tried to counter negative backwash by basing more oftheir material on authentic sources, and reducing the number ofdecontextualized sentences. Discrete item testing (knowledge ofindividual points of language) is balanced with global testing (successfuluse of more than one language skill in more extensive chunks of text). Inaddition, the testing of speaking has become a more important feature,and has been made more natural and communicative. It is also refreshingto see some examining bodies insist on the use of dictionaries in theexamination room: in real life, students would not be isolated from suchuseful resources, so why should the examination not allow this? However,the backwash effect remains predominantly negative and encourages amodel of learning summed up in the left-hand column of Table 1.(Broadly speaking, the two approaches described correspond to left andright brain learning—see, for example, Gerngross and Puchta, 1992.

I would like now to discuss in more detail some of the characteristicslisted in Table 1.

Tests are designed to discriminate proficiency, progress, and achievement.Indeed, some tests would be regarded as inefficient if all candidatesenjoyed equal success. This is innocent enough, and, in administrativeterms, very useful. It does, however, tend to encourage a view of studentsas 'good' or 'bad', 'strong' or 'weak'. Such a classification may be thefirst step towards a fatalism that assumes some students are born to failand others are 'natural' language learners. The victim of this Manichaeanview of the classroom world is usually the so-called 'bad' learner, who is

18 Luke Prodromou



ownloaded from


Table 1 Testing Teaching

exercises (multiple-choice, etc.]failureweaknesserrorphobiamarksfearanxietyteacher controltextbook inputjudgementmistrustindividualism, competitionimpersonalityinsensitivityisolated sentencesfragments of textformculture-boundtext+questionssolemnityboredomextrinsic motivationproduct

taskssuccessstrengthlearning from errorachievementconfidencepleasurelearner independencelearner inputsupport (from teacher and peer group)rapportthe group, co-operationpersonalizationsensitivity to learnerstextwhole textscontentculture-sensitivelead-in, follow-uphumourinterestintrinsic motivationprocess

condemned to failure by the pressures of examination preparation.Testing is a straitjacket that cramps personal learning styles anddiscourages the 'weak' learner's potential for growth.

Error In testing, there is a pre-occupation with accuracy and error that reachesalmost behaviourist proportions. Sadly, 'errormania' on the part of theteacher encourages 'errorphobia' in the learners: a reluctance by 'weaker'members of the group in particular to take risks for fear of makingmistakes, losing marks, and thus slipping down in the classroomhierarchy. The result is a silence arising not from the learners' ability butfrom discriminatory classroom procedures. Research into the goodlanguage learner has highlighted the fact that these learners are risk-takerswho are also able to learn from their mistakes (Rubin 1987, Wenden 1987,Stern 1983). By not encouraging learners to learn from their mistakes andwork out the rules of the language for themselves, conventional testingprevents the full development of the cognitive aspects of learning, therebycontradicting what we have come to consider as good teaching practice.

Marks versusachievement

A classroom climate dominated by testing will give students theimpression that what matters in language learning is the mark they get, notonly in tests, but also for classroom performance, assignments, andhomework, even though these may have no direct connection with the finalexamination. The students' frequent demand for feedback on all of theirerrors and their preoccupation with their weaknesses rather than theirstrengths have their numerical equivalent in the pursuit of high marks orgrades. Although the view that language acquisition is easily quantifiablemay encourage students to work harder (extrinsic motivation), it alsoobscures the importance of concepts which are not always easy tomeasure, such as appropriacy, quality, and attitude in learning.




ownloaded from


Anxiety versuspleasure

Textbook inputversus learner

input

The discourse ofexamination texts

Tests and examinations are closely associated in learners' minds withanxiety; it is doubtful whether performance, even in tests, is facilitated byan attack of fear and nervous tension but, in educational terms, these are amajor obstacle to learning. Most recent approaches to language learningwould accept the importance of affective factors in the classroom. Thefeatures of orthodox testing I have described so far all contribute towardsraising the learner's affective filter, and thus placing barriers in the way ofefficient learning. Moreover, a great strain is placed on classroomrelationships when the teacher is called upon to play the role of thestudents' judge and executioner: when testing comes through the door,rapport between teacher and learner often goes out the window.

Anxiety about covering the examination syllabus means teachers areafraid to take risks with material not manifestly related to theexamination; students may also become impatient with material whichdoes not seem to be in the form of examination practice. This has multipleconsequences: textbook and teacher input are the order of the day, and thematerial chosen may be irrelevant to learners' personal needs or culture;even when the material is potentially interesting it is not taught for contentbut for form, as this serves the narrow requirements of examinationpreparation.

One-dimensional or anaemic textbook/teacher input is to some extentinevitable when the public examination is also an internationalexamination, available in countries with widely differing cultures.Challenging and culturally relevant material is watered down on theprinciple of the lowest common denominator. Examination material willtend to reflect the culture in which English is spoken as a first language.The problem of alien and alienating content is a parallel one to that of the'global textbook', but in an even more acute form. The remedy might be agreater use of local and learner input, but this is an option rarely adoptedby teachers and students straitjacketed by examination syllabuses andmaterials: material that will not 'come up in the exams' will rarely be'brought up' in class.

The content of texts used in examination preparation is not onlyimpersonal and culture-bound, it is often a peculiar variety of English,which is neither fact not fiction. Test items like the following are unique tothe examination genre—they are literally context-less and content-less,about nothing and nobody in particular:

1 I'd like to visit India more than any other country in the world.India is

2 The flight to Moscow lasted three and a half hours.It took

Pronouns in English usually refer to somebody or something previouslymentioned, but the pronoun in sentence 1 refers to neither: it is pure form.We do not know who the T referred to is, and we will never know why heor she would like to visit India.

20 Luke Prodromou



ownloaded from


The definite article in sentence 2 is tantalizingly specific, but in fact thecohesion is misleading: there is no flight, definite or indefinite; there is nocontext to which this text belongs and to which we can refer if we want toknow more. It is difficult (though not impossible, as I will show later) torelate these propositions to a linguistic or real-world context.Functionally, we do not know whether these texts are parts of a narrativeor argumentative text.

Solemnity versushumour Humour, like context and content, is considered inappropriate in testing.

Tests cannot be funny because there is not thought to be room for humourin the solemn ritual of the examination process. Thus, testing becomes apretext for numerous practices which in any other pedagogic context wewould reject out of hand as inimical to good language learning.

Product andprocess The two methods of imparting language that I have referred to as testing

and teaching differ in one overarching essential: the first focuses almostexclusively on the product (the language to be taught), while the secondaims to make the process of imparting language both interesting andfulfilling. In the former, the learner's potential, both linguistic andpersonal, is downgraded; in the latter, it is encouraged. The obsessionwith the linguistic product, and the sacrifice of rich pedagogic processes,is usually accompanied by cries of 'We don't have time', 'I must finish thebook', 'Cover the syllabus', etc. Preoccupation with the end-productobscures the importance in language teaching of two factors mentionedearlier: classroom management and rapport. If the examination syllabusand its accompanying exercise types are the destination for manyteachers, the process by which they reach that destination is the journey: Ihave been arguing that this journey, which depends so much on goodmanagement and rapport, should be both enjoyable and educationallysatisfying.

Transferringtesting

procedures toteaching

procedures

Tests and examinations are not going to go away; the product, bothlinguistic and commercial, will continue to be packaged, marketed, andsold. Any suggestions one makes concerning the transfer from testingprocedures to teaching procedures must take this fact of educational lifeinto account. The examples I give in the final part of this paper are basedon exercise types commonly used in examinations. I will try to show howthe most unpromising testing material might be made into a morechallenging vehicle of personal expression, without the teacher having toabandon the book or the syllabus.

The overall principle behind the techniques described below is the shiftfrom teacher control to student control. The tasks will therefore involvethe use of learner input, which I feel is a key element in transformingnegative into positive backwash and in making examination preparationmore of an educational activity than it is at present in most examinationclasses.




ownloaded from


1 The examiner's hat

(i) Students complete sentence-level multiple-choice, gap-filling, ortransformation exercises in the conventional way.

(ii) Students rewrite the test items to reflect their personal views, usingthe textbook or testbook as a guide. Thus, if an original sentence inthe test says 'Stamp collecting is the most enjoyable hobby I know',students can replace either the subject or the adjective with items oftheir choice. Students may also replace impersonal or 'non-existent'subjects such as 'he', 'she', or 'John', with the names of friends,people in the class, members of their family, or famous people, ormake the sentence interesting or amusing in any way they wish. Theobjective in each case is to make the utterly forgettable originalsentence become memorable in some way.

(iii) In groups, the students re-cast their personalized sentences in thestyle of the test format they are working on (multiple-choice, gap-filling, transformation exercises, etc).

(iv) The groups swap their personalized test items and do each other'stests.

(v) The teacher checks and gets feedback on the form and in particularon the content of the sentences.

2 Gender bending

(i) Get students to rewrite sentence-level multiple-choice andtransformation exercises by asking them to change all femininesubjects into masculine and vice versa. The results will be bothsurprising and memorable.

(ii) Focus on content: consider whether the resulting sentences are (a)correct (b) acceptable. A lively debate will invariably ensue as, forinstance, when 'He's such a naughty boy; it's amazing what hismother lets him get away with' becomes 'She's such a naughty girl;it's amazing what her father lets her get away with' or when 'Havinglaid the table, Mrs Jones called the family to supper' becomes 'Havinglaid the table, Mr Jones called the family to supper.'

(iii) Ask students to summarize the results of this exercise into a chartsimilar to Table 2. (This is a small sample based on an authentic,international public examination!).

(iv) Ask students to write an argumentative composition on equality ofthe sexes based on the 'data' given in Table 2.

3 Transformation tennis

The following exercise breaks down the pattern of serried ranks ofstudents—heads in books, no eye contact—by using the classroom spaceas a kind of verbal tennis court.

22 Luke Prodromou



ownloaded from


Table 2 Males Females

stay out late have to come home earlycan push open doors make chocolate cakeearn a lot of money give up their jobs to look after childrendrive quickly and carelessly look carefully before crossing the roadmake difficult decisions are afraid to go into the seatake an active part in politics tell liesbecome priests become nurses

(i) The class form into two teams. A student in team X 'serves' the firsthalf of a pair of transformation sentences, for example by saying'India is the country I would like to visit more than any other.'

(ii) A student from team Y serves the sentence back in the form of anappropriate transformation, using the stem provided in the book: Thecountry (I would like to visit more than any other is India). Team Yscores if the transformation is correct.

(iii) The teams take it in turns to serve.

This technique can be applied to any textbook exercise which is in twodistinct parts: matching, multiple-choice, word-building (e.g. noun toadjective).

4 Connecting the fragmentsThe following exercise takes impersonal fragments of language withminimal content and gets students to make them into an integrated part ofa whole text.

(i) Complete a sentence-level test in your usual way (multiple-choice,transformations, word-building, expansion from notes to completesentences, etc.).

(ii) Ask students to write a composition to include at least one of thepractice sentences. They can incorporate the sentences at thebeginning, middle, or end of their composition.

(iii) Students circulate their compositions round the class (or you canstick them up on the class noticeboard as an 'exhibition'). Studentsread each other's texts and try to identify and make a note of thesentences from the original textbook exercise.

This exercise not only provides creative composition practice, but alsorevises those sentences which, no sooner practised, are usually forgotten.

Management My suggestions may seem to imply that the problem of negativetechniques backwash is one of course design and methodology. It is, however,

especially in its covert forms, chiefly a problem of attitude and rapport.Teachers express attitudes towards learning not only in their choice ofmaterials and methods, but also in their approach to classroommanagement. This discipline—which involves the use of time, space,voice, and gesture—weaves subtle messages which can motivate ordemotivate a class. For this reason I would like to end with a brief




ownloaded from


checklist of management tactics which will tend to mitigate some of thefeatures of testing, and encourage an ideology of co-operative learning.

From multiple 1 When you ask the class a question, allow the 'weaker' students somechoice to personal thinking time—do not make question and answer routine a race to the

choice right answer.

2 Do not stand too close to the student who is answering the question,thereby excluding the rest of the class: use space and distance to createan inclusive, group feeling.

3 When you get a 'right' answer do not just move on to the next item:ask other students 'Do you agree?', 'What have you got?' Do notreveal the right answer too soon. The process is as important as theproduct.

4 Give students time to look at questions before they listen or read—thiswill make the task more directed, and help develop skills rather thanmerely test them.

5 Encourage students to share—establish the idea of tests as a groupactivity alongside individual testing tactics.

6 Ensure smooth linking of the stages of the lesson—avoid thedisconnected, random fragments, the rag-bag so characteristic of testmaterial and examination preparation classes.

7 Avoid saying things like 'work quickly', or 'you've got one minute todo this—hurry up'.

8 Arrange desks in such a way that students can see each other and makeeye contact.

9 When you ask a question or discuss a problem use eye contact toinclude the whole class, not just the 'best' students.

10 Use your voice to suggest that error is a useful contribution to theclass, not an unfortunate lapse on the part of the student. Try a fall-riseintonation ('Yes, but . . .') rather than a fall ('No').

Received December 1993

24 Luke Prodromou



ownloaded from


ReferencesAlderson, C. and D. Wall. 1993. 'Does washback

exist?' Applied Linguistics 14/2: 115-29.Fabian, P. 1982. 'Examinations: why tolerate their

paternalism?' in J. B. Heaton (ed.) LanguageTesting. London: Modern English Publications.

Gerngross, G. and H. Puchta. 1992. CreativeGrammar Practice. London: Longman.

Harmer, J. 1991. The Practice of English LanguageTeaching. London: Longman.

Heaton, J. D. 1990. Writing English Language Tests.London: Longman.

Holt, J. 1964. How Children Fail. Harmondsworth:Penguin.

Howatt, A. P. R. 1984. A History of EnglishLanguage Teaching. Oxford: Oxford UniversityPress.

Hughes, A. 1989. Testing for Language Teachers.Cambridge: Cambridge University Press.

Peck, A. 1988. Language Teachers at Work. London:Prentice Hall.

Richards, J., J. Platt, and H. Weber. 1985. TheLongman Dictionary of Applied Linguistics.London: Longman.

Richards, J. C. 1990. The Language TeachingMatrix. Cambridge: Cambridge University Press.

Rubin, J. 1987. 'What the 'good language learner'can teach us'. TESOL Quarterly 9/1:41-51.

Seaton, B. 1982. A Handbook of English LanguageTeaching Terms and Practice. London: Macmillan.

Stern, H. H. 1983. Fundamental Concepts ofLanguage Teaching. Oxford: Oxford UniversityPress.

Wenden, A. 1987. 'How to be a good languagelearner' in A. Wenden and J. Rubin (eds.) LearnerStrategies in Language Learning. EnglewoodCliffs, N.J.: Prentice Hall.

The authorLuke Prodromou works for The British Council inGreece. He has been involved in training teachers forthe DTEFLA and DOTE and is an assessor/moderatorfor these schemes. He has also been a member of theUCLES CTEFLA Scheme Committee and the teamworking on the new Cambridge Integrated LanguageTraining Schemes (CILTS). He is the author of MixedAbility Classes (Macmillan) and several textbooks forexamination classes.




ownloaded from


Backwash Effect

Documents

cambridge

examination

good language

good language

examination

public examination

examination

imparting