See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228390033

Language Testing: Then and Now

Article · July 2010

DOI: 10.3126/nelta.v8i1.3379

CITATION READS

1 9,666

1 author:

Ashish Giri
KIIT University
1 PUBLICATION   1 CITATION   

SEE PROFILE

All content following this page was uploaded by Ashish Giri on 23 March 2014.

The user has requested enhancement of the downloaded file.

 Language Testing: Then and Now/49

LANGUAGE TESTING: THEN AND NOW

- Ram Ashish Giri

Department of English Education

Tribhuvan Universtiy, Npeal

Abstract
The article is a brief historical overview of English language testing, particularly the testing of
English as a second or foreign language. It offers a discussion of how dffirent language testing
trends have emerged out of the changes taken place in the field of linguistics, particularly,
applied linguistics and teaching of English as a second or foreign language. In discussing the
trends, the British, North American and Australian contexts, in particular, hqve been
considered. Finally, some issues of the present day language testing are reviewed.

1. Introduction
'Tests' and 'examinations' are ancient practices. Their origin, oflen subject to
interpretations, can be traced back to the 'pre-historic' period of education. In the Indian
sub-continent, for example, public discourses and disputations or demonstrations of
acquired abilities in a given area were a common practice in the'ancient academic
period. The Greeks and Romans are also said to have practised some form of academic
examinations in their glorious past. About four thousand years ago, the Chinese are
believed to have adopted 'examination' in an elaborate form of a measure of ability for
the first time. However, the form of examination that is current in academic settings
today has evolved from and is a development of the examination that was in practice in
the 19tn century. Today, the word'examination'means a series of systematic'tests'of
knowledge, skill or of special ability to be carried out by an individual or an authority.
So far as language testing is concerned, it is generally assumed that the history
of English language testing is as old as the history of English language teaching itself
because testing has always been an integral part of any English language teaching
(ELT) programme, which probably began in the 15th century with the ordinance
promulgated by Henry the 5'n that English should be adopted as the language of royal
correspondence in the place of French. This royal ordinance is believed to have
facilitated the development of English language teaching methods, writing of teaching
and learning materials and designing language testing strategies. In the beginning, there
was a serious lack of teaching and leaming materials. So, methods to be adopted,
materials to be used and testing strategies to be followed largely depended on the
concerned tutors. As a result, the Engli,sh language teaching and testing did not flourish
much. In fact, it was only in the 16tn century when serious attempts were made to
produce a scholarly description of the English language thereby providing a foundation
for its teaching and learning. This has been marked as the formal beginning of teaching
English as a foreign language which later was also termed as English as a second
language depending on where the teaching of the language took place.

Journal ofNELTA Vol. 8, No. I-2 December, 2003

50 / Ram Ashish Giri

Teaching English as second/foreign language was commonplace in the l Tth

century due to leligious ad political upheaval which had ,brought a large number of
foreigners to Brita;, but comparatively less so in the 18th century as the number of
immigrants dramatically dropped following truce in politics. Outside Britain, France
and Denmark were the first countries where teaching of English as a foreign language
started. Learners in these countries had to depend on French as the medium of
instruction because teaching and learning materials did not exist in any other languages.
The teaching of the Engliih language entered a new phase with the publication of
Johann Christian Fick's 'Practical English Course in 1793, the earliest gratnmar
translation course. However, in the third world countries, the teaching and learning of
English started in 1797 with the publication of John Miller's The Tutor'
English language testing in England took a significant tum in the 1850s when a
system of public examinations was established which was controlled by the universities.

Purpose of the article

The purpose of the article is to offer a brief historical overview of English
language testing, particularly testing English as a second or foreign language. The
griiisn, North American and Australian contexts, in particular, have been considered.
The paper also discusses the trends in language testing that emerged out of the
develtpment of linguistics, particularly applied linguistics and teaching English as a
secondTforeign language. Finally, some issues in the present day language testing are
reviewed.
Language testing is an integral part of language teaching and ' both testing and
teaching ur. rcL clearly interrelated that it is virtually impossible to work in either field
withouf being constantly concerned with the other' (Heaton, 1988.viii). Also, trends in
language testlng have tended to follow trends in language teaching which in turn have
followed trends in linguistics, pafticularly applied linguistics. Therefore, trends in
language testing have been analysed here in terms of the development of a linguistic
theory, its subsequent application to ELT and its influence in the emergence of English
language teaching and testing approaches and methods'
The article is limited to the findings of a preliminary review of some of the
literature in the field of language testing, particularly second or foreign language testing.
The author does not advocate for or against any particular trend, nor does he make a
case for a particular approach to language testing. He simply makes a descriptive
analysis of the history of testing English as a second or foreign language. The readers of
the article should note that there is a difference between Engiish as a second language
and English as a foreign language. The writer uses these terms interchangeably though
he uses them to refer to specific context for English language teaching and testing
occasionally.
There are three sections of the article. The first part is a brief introduction, and a
survey of the British, North American, and Australian contexts and the development of
the testing English as a second or foreign language. The second section describes
 Language Testing: Then and Now/51

different trends in language testing associated with the development periods of

linguistic theories and their influence on English language teaching and testing' The
thiid section analyses the emergence of various language testing purposes and presents a
paradigm that governs the ESL/EFL testing today. A brief reference to the present day
bnglish test ofihe School Leaving Certificate (SLC) Examination has been made in the
final section of the article.

2. Origin of Testing of English as a Second or Foreign Language

2.1 The British Tests

Testing of English for foreigners started much earlier in Britain than in America. In fact
the Un]versit/of Cambridgi became involved in overseas testing within a few years of
the establishment of the public examinations. The university sent out examination
papers overseas for the firit time in 1963 to examine ten candidates and continued to
develop its overseas examinations in the next four decades of the 19th century. By 1898'
the University of Cambridge had 36 colonial centres with 1220 candidates. However, its
formal entry in testing the English of foreigners was not until 1913 when it instituted the
examination for the bertificate of Proficiency in English meant to be for "foreigners"
who sought proof of their practical knowledge of the language with a view to teaching it
in foreign schools.
i.eflecting the growing interest in the direct method of teaching, which required
of teachers' reliable rand of the language for active classroom use rather than
"o-t
academic or descriptive ability, the certificate examination was, however, modelled on
the traditional native speaker language syllabus, with the essay as the key feature. In
addition to a paper in phonetics, there were a literature paper and a paper involving
translation. Oiaf examinations, writing questions such as letter writing and composition
either on a narrative, descriptive or an imaginative topic and reading and writing main
points, and a summary of a story were added in the certificate examination in the
iubsequent years. The- certificate received a major boost when in 1931 , the University
recognised it as the equivalent of the standard of English required of all.students, British
or f6reign, before tlieir entrance to the University. With the growing demands of
certifrcaiion in English, different levels of the certificate were introduced and later
conducted through the British Council.
For the Cambridge tests, the issue of reliability was not an important
consideration, and so the examinations remained'untouched by psychometric notions.
By comparing the College Board's examination of 1931 with the one practised until
recently, one can easily see the lack of concern about reliability. While the Board's
examination had its concentration on language with true-false questions without
curricular concern, the Cambridge examination put emphasis on curriculum with
literature and absolute reliance on subjective grading. Therefore, it can be concluded
that the Cambridge examinations, which were innovative and almost non-academic in
their desire to test language use rather than knowledge about language, are more
concerned about what to test (i.e., about curriculum) than about how to test.
The ESL/EFL tests of General Certificate of secondary Education (GCSE) are
produced and administered for the non-English speaking background (I'JESB) students

Journal ofNELTA Vol. 8, No. l-2 December. 2003

52 / Ram Ashish Giri

enrolled or seeking effolment in the British schools. At the higher level, examination
boards develop and administer O (ordinary) and A (Advanced) level examinations and
also ESL and EFL proficiency tests. The EFL and ESL tests are given to applicants to
the British instituti,ons to determine their ESL/EFL proficiency. The International
English Language Testing System (IELTS), which is for the students of Year 11
onward, is a measu.e oi English Language proficiency for the students seeking
admission for higher education or training in British or Australian universities, and has
been jointly developed by University of Cambridge Language Examination Syndicate
(UCE"ES), The British Council (BC) and International Development Program (IDP) of
Australian Universities. The test reflects the ideas of communicative language teaching
and is probably the first standard communicative language test administered over a large
population. tbl,ts, which is widely welcomed by the British and Australian
universities, consists of two sections - General and Modular. The general section
comprises of a listening test and an oral interview intended to test the oral skills' The
modular section, on the other hand, is intended to test the written skills, reading and
writing. The modules are further divided into sub-modules.
The overall format for the modules are the same; they all contain texts from
books, journals, reports, etc., related to a specific subject area and involve candidates in-
study skills necessary for academic studies.
British tests are highly innovative in content and format, and lay great emphasis
on validity on the examination construction procedures, which rely on expert
judgement.

2.2T|r.e North American Tests

The Immigration Act of 1924 allowed visas outside the quotas for foreign students, but
the Immigration Department was unhappy that many non-quota immigrant students
gained aJmission to ttre United States totally unfit because of their insufficient
knowledge of the English language. The Deparlment asked for an indication of the exact
knowledge of the English language a student must have before he or she can be
accepted in the United States of America (Spolsky 1993:4).
th. Coll"ge Entrance Examination Board was asked to develop a test that measured
immigrant students' knowledge of English and determined their ability in the language.
The American Association of College Registrars also requested the Board to consider an
addition to their special examination designed to test the ability of a foreign student in
the use of the English language, which later became a mandatory requirement for
attendance by American College institutions. Consequently, the Board set up a
commission, which recommended an examination to test ability to understand, to read
English intelligibly, and to understand spoken English. The aural comprehension test
should include simple English prose read slowly, and simple directions given
conversationally. The Board put forward, in 1930, the English Competence
Examination, which is perhaps the earliest test intended specifically for English as a
Second Language in the Unites States. The specifications of the examinations were as
follows:

Journal ofNELTA Vol.8, No. l-2 December, 2003

 Language Testing: Then and Now/53

1. Four one-paragraphpassages
- One narrative
- One historical, topical or journalistic,
- One critical, and
- One scientific
2. A longer passage dealing with debatable ideas,
3. A direct dictation, and the reproduction from memory of a dictated passage,
4. An oral test with ten topics prepared for the examiner,
5. Composition on a selected topic.
In the last two decades, students seeking eruolment in the North American
universities have been given one of the three tests, Comprehensive English Language
Test (CELT), Michigan English Language Assessment Battery (MELAB) or Test of
English as a Foreign Language (TOEFL). Among these, TOEFL is considered to be a
secured evidence of English language proficiency by most North American universities.
These tests have traditionally involved the assessment of listening, reading, vocabulary
and grammar. The TOEFL has also included Test of Written English (TWE) and Test of
Spoken English (TSE) as direct measures of oral and writing proficiency. More
recently, the TOEFL has been computerised. Students can enter their answers directly
on the computer and results are known within days. The test called American Language
Institute of Georgetown University (ALIGU) test is another test of academic English
given to applicants of scholarships awarded by various US government agencies.
For the school level ESL/EFL proficiency, only the Language Assessment
Battery GAB), which assesses receptive as well as productive skills, and Secondary
Level English Proficiency (SLEP), which assesses receptive skills, are given.

2.3 The Australian Tests

The Australian system of education follows the British one and the tests of ESL
or EFL are more or less similar to those of the British. Institutes like Australian Council
for Educational Research (ACER), Language Testing Rqsearch Centre (LTRC) and
National Centre for English Language Teaching and Research (NCELTR) have
developed and implemented test of English language proficiency for specific levels,
tasks and situations almost all of which are for the students of Non-English Speaking
Background O{ESB) studying in ESL programmes in Australia. More recently,
Australian Second Language Proficiency Ratings (ASLPRXIngram 1984), a structured
interview procedure and rating scale has been designed to measure general proficiency
in a language learned as a second or foreign language. Candidates are rated on a
descriptive (criterion-referenced) scale in which the language proficiency in the four
basic communicative skills are specified by a number, title and behavioural description.
The scale consists of descriptions of language behaviour at nine levels - 0 to 5, five
being the native-like proficiency. The ASLPR provides three kinds of information about
each level - (1) a description of the language behaviour appropriate to the level, 2) a
series of examples of observed behaviour and (3) a series of comments explaining the
keys features in a given area.

Journal ofNELTA Vol. B, No. 1-2 December, 2003

54 / Ram Ashish Giri

for use
The ASLpR is based on the absolute proflciency ratings and is_designed
pemlits
with learners whose education and employrnent is more diverse. It, therefore,
;i;; i"g*e of differentiation. The ASi-PR is not a test instrument, rather it is a scale
Language Roundtable (ILR)
una u se"t of procedures which are based on Interagency
Testing Foreign Languages/
Oral Interview derived from the American Council on the
rater has the
Educational Testing r"*ir. (ACTEFL/ETS)(Alderqol 9t aI. 1987)- The
more
lit..ry 16.eplace tli suggested tasks or exercises with the ones he or she regards level and
uppiJp.i"t.,'provided ttr"aT ttre new tasks and exercises are of equal diff,rculty
linguistic or communicative levels
complexity. However, no guidance at the functional,
complexity is provided'
as to how to make the taski or exercises of equal difficulty
and
To sum up this section, thousands oi students/individuals, in every part of the
their
world, take various English language proficiency tests each year to- demonstrate
are then used
ability/proficiency ir-gigiirft as"a f6reign language, the s.cores of which
Uy aiffi.*t institJions"of English Telkine-countries^like USA, UK, Australia and
purposes such as offering
i*iau,fo. ,"r."nirr! tn. .*Aiiates for u tt.tl*btt of different in career' Test
admission to an edu"cational programme, employment or advancement
scores are related to varioui aspects of pioficiency demonstrating
the candidates'
reading and writing' In the last
iung,rug. ability in different skiun listening speaking,
developed to measure different aspects
thrJe d"ecad"s, n,r-.rous test batteries have 6een
of English language proficiency of the non-native speakers.
TOEFL
American engUsfr language test batteries such as MELAB, ALIGU and
have been in use in Xsian cointr]es including Nepal for a long time
now. IELTS made
its way to Nepal about 10 Years ago.
"Ingeneral,Britis'handQ.{orth)AmericanEFLproficiencytestsrepresent
different aiproaches to language
"on
test development. North American tradition in
il&;;g; i6rt ir heavily baied properlies of tests' Issues such as
_psycho*.tti9
i"ri"uuifitv una and prediciive validity ar-e of particular interest in this
oUj".ti"ity of scoring and generalisibility of results play a dominant
tradition. Hence"o"turrent
often
.ot. in the development of test metho?s. Foiexample,-multiple-choice items aretest
used in testing r.L.pii* skills to gain desired internal consistency even if-the is
expected to measure communicative competence. Moreover, in orde-r to achieve high
in
inter-rater reliability, ih; ;;" of trained i"or.rr and detailed specific instructions
.onaurting interviJws are highly recommended for testing productive skills in the
tradition.
The British tradition puts emphasis on the specification of test content and
content and
expert judgement. While reliaUility reieive less attention in this tradition,
face vaiidity are the major of the test designers' It can be generalised that
"orrceirrs
British tests enjoy ;; variability in their formats and include various communicative
activities.

3. Trends in Language Testing

"in "second
Modelling oi foreign language testing is not a. recent concept.
Theorists and laireuage teaching meth6dologists have developed. and used
language-
point
t"rffi,"ai"ir, *rr"i"ii--urcn"a tte language tEaching models existing-in_particular
of time. In others words trends in ling'uage testing have always followed trends in
language teaching.
Vol. B. No. I-2 December, 2003
Journal ofNELTA
 Language Testing: Then and Now/55

Language testing specialistsi theorists have always stressed the need to base the
development and use of language tests on a theory of language proficiency. More
recently, they have called for the incorporation of a theoretical framework of what
language proficiency is with the methods and technology involved in measuring it.
Bachman (1990:9) describes the complexity of the problem of defining and measuring
ability in the following terms:
All language tests must be based on a clear definition of language abilities, whether
this derives from a language teaching syllabus or a general theory of language
ability, and must utilise some procedure for eliciting language performance.
However, there are different ways of looking at language ability, and as a result, there
is a wide variation in its definition. These different definitions of language ability
focus on different aspects of the ability, and so developing a single language testing
design to accommodate all of these aspects is very complicated. Furthermore, the test
instruments or methods used to elicit language abilities are themselves based on the
assumed level of the abilities, making it unceftain whether the test measures will elicit
data to characterise an individual's language performance in non-testing situations.
Language testers, therefore, face a complicated dilemma of attempting to measure
abilities that are not precisely defined, and of using methods to elicit language abilities
that themselves are manifestations of the very same language abilities. In other words,
the test methods or data elicitation procedures one uses to measure language abilities
are characteristically related to the way one views these abilities. The most important
task of a language tester, therefore, is to define language abilities in such a way that
the test methods or response elicitation procedures applied to elicit language test
performance that is characteristic of language performance in non-testing situations
(Bachman 1990:9). Given below, then, is an account of different trends in language
testing and the ways the language has been viewed in these trends.

3.1 The Pre-Scientific Trend

Language testing cannot be separated from an understanding ofthe nature oflanguage,
of language abilities, and therefore, of its learning and teaching. Different theorists of
language testing, therefore, viewed language and language abilities differently, which in
consequence, have emerged as language testing trends in different but sometimes
overlapping periods. Prior to 1960s, for example, language was viewed as a means to
approach the target culture and language learning as a way of training the mind
(McGarrell 1981). The target language was an object of study, which could be separated
from its context for the purpose of teaching and learning. In this 'separatist' view
(Davies 1968:216), the relationship between society and language is not considered
important. The target language is separated from its use for the purpose of encouraging
learners to an exhaustive study of the target literature, and for training their mind
through practice in dealing with varied learning situations and materials. Language
leaming, therefore, was viewed as a process of learning grammatical rules and
vocabulary and applying those rules and words in writing and translation exercises.
Language tesling was seen as testing the learning of words and grammatical accuracy
through writihg exercises.

Journal ofNELTA VoL B. No. I-2 December, 2003

56 / Ram Ashish Giri

This period, in the history of language testing, is considered to be the 'intuitive'

stage (Madsin 1983) or the pre-scientific era (Spolsky 1978) in that decision about
teaitring and testing mainly depended on the personal discretion or subjective
judgement of the teachers or testers.
The pre-scientific trend embraced the traditional grammar - translation approach
in which learners were required to apply the rules of grammar taught to them
deductively and the vocabulaiy they learned by memorisation. The grammar translation
course was basically a grammatical progression, which included grammatical structures,
and vocabulary need.i to .*pr.s thim, both selected arbitrarily on the basis of
principles of easy to difficult, familiar to unfamiliar and concrete to abstract' The items
that were viewed as simple, easy and familiar were taught before the ones that were less
frequent, abstract, complex or difficult'
The merit of this trend of language testing is that it allowed a global evaluation
of the learners' ability in the targeilanguage tfirough their composition and writing
activities which required learneri to synthesise their knowledge of the rules and
vogabulary and to produce syntactically acceptable utterances. To help learners
gain
coiitrol over formal accuracy, tasks sr,.it ut an analysis of sentence structure, labelling
its parts or synthesising parls.into larger units were commonly practised (Ingram 1985)'
To reinforce the mastery o,r.r grammatical accuracy further, abundant writing activities
such as translation, letter, prdcis, and open-ended answers based on reading and
"rruy,
cbmprehension were organised.
'ihe testers made their judgement about a. candidate's
abiliiy in the target langirage on the basis of the knowledge of grammar and vocabulary
exemplified in these forms of writing (Madsen 1983)'
This trend is characterised by a lack of testers' concern for statistic analysis,
objectivity, or test reliability (Howard 1980). In this app_roach, the onus was placed on
ifr. r"US.liive judgement of the learners' language proficiency Ther9f91e,.a language
test within the framework of this trend was tfie responsibility of the individual
teacher
who set his/her own standards or criteria. A typical language test comprised of
translation, framing of question answel and composition exercises. Oral tests were
rarely given.

3.2 The Psychometric-structuralist trend

During 1960s, the behaviourist language learning theories_inspired rytlly by_Skinner's
work it 957), and structural linguiitici emerging out of -the work of Fries (1952) and
Bloomfield (tl::;, evolved into what Spolsky (r978) calls_ a psychometric- structuralist
trend. In this trend, which is also sometimes known as the era of scientific language
testing (Madsen 1983), language was viewed as a multitude of discrete-point
patterns
which could be learned Uy t[e stimulus-response habit formation process (Ingram
1985). Language learning was "... a process of^acquiring conscious.control
of the
phonotrogicil, g-rammaticil, and lexical patterns of a.second language, largely through
rt,tay' ani anulysis of these patterns as a body of knowledge" (Carroll 1983).
of the
Severai approachesio language teaching appeared during this period some
notable ones are - the Audioling-uaiapproach, tie mimicry-memorisation method,
the
The structural approaches to
oral-aural method, the oral struclural method, and so on.
Language teaching regard language as a set of structures, and language learning as

Vol. B, No. l-2 December, 2003

Journal ofNELTA
 Language Testing: Then and Now/57

gaining mastery over the structure through the process of repetition and practice.
Language ability is seen as the ability to handle discrete elements of the language
system and develop aspects of individual language skills.
The discrete-point approach to testing measured language proficiency by testing
learners' knowledge of the discrete items of grammar and vocabulary and of discrete
items of language skills by taking one item or aspect at a time.
Discrete-point tests became the most widely used tests worldwide in 1960s and
1970s and are still popularly practised in many parts of the world. So much so that the
psychometric-structuralist trend in testing a second or foreign language became
synonymous to objective testing. Lado's work (1961), which has been seminal in
introducing objective testing to second or foreign language testing, claimed a universal
appeal and received a great deal of support from the linguists, methodologists, teachers,
course designers and test developers everywhere in the subsequent years. It "has
correctly pointed to the desirability of testing for very specific items of language
knowledge and skills, judiciously sampled from the usually enoffnous pool of possible
items. This procedure makes it a highly reliable and valid testing" (Spolsky i978 quoted
in McGarrell 1981).
There has been a lot of criticism of this approach to language teaching and testing
also. Ingram (1985) suggests that language is not just the sum of its parts, butthe parts
have to bb.mobilised and integrated together to carry out particular tasks in particular
situations. Davies (1963) rejects this approach by saying that whole language
development cannot and should be equated with the separate development of its
constituent parts because the whole is always greater than any one of its constituent
parts.
Oller (1979) makes a similar comment in his seminal work in which he discusses the
defect of the discrete point approach to language testing by saying that it suffers from
several deficiencies, and suggests that the problem with this approach is that it depends
on language proficiency being neatly quantifiable. He outlines its deficiency in the
following terms:
Discrete point analysis necessarily breaks the elements of language apart
and tries to teach them (or test them) separately with little or no attention to
the way those elements interact in a larger context of communication. What
makes it ineffective as a basis for teaching or testing languages is that
crucial properties of language are lost when its elements are separated. The
fact is that in any system where the parts interact to produce properties and
qualities that do not exist in the part separately, the whole is greater than the
sum of its parts...organisational constraints themselves become crucial
properlies of the system which simply cannot be found in the parts
Separately (Oller 197 9 :I 12)

As the quotation above implies, language and its social context are
complementary to one another. Language knowledge, therefore, must be tested in
language contexts to see if a person can communicate appropriately in the given context
of situation. Testing formal knowledge of language, i.e. linguistic competence, is
necessary but not sufficient to predict that the person can use the language effectively in

Jourrql of'NELTA Vol. 8, No. I-2 December, 2003

58 / Ram Ashish Giri

a given situation. A person taking a driving test or music test, for example' must
demonstrate their driving or perfoining abiliiy as a whole. The knowledge
of engine
parts or the keyboard does not necessarily make him or her a good driver or
performer'
weir (1988) goes on to suggest that grammatical (linguistic) competence is not a good
in
indicator of one's communlJative skil'is at all. A tester, therefore, should be interested
the development and measurement of communicative competence- rather than of
linguistic cbmpetence. Weir's suggestion is in line with Spolsky (1978) and Morrow's
that, instead of attempting l9
ltilOy commints which *.t. -id. a decade earlier should
,rtuUtir6 a person's knowledge of a language in terms of elements and skills, one
setting.
attempt to test that person'r ubility to perform in a specified sociolinguistic

3.3 The Psycholinguistic-sociolinguistic trend

A learner ii exposEd to the targef language through its skills (listening, speaking,
reading and writing) and elements (sounds, words, grammar, etc.). However
a language
is learned in its eniirety, i.e. language skills and elements are leamed as related
parts,
not as isolated component, oiu iystem. This holistic or integrative approach to
is unitary process
language and languag. t.*tting implils that learning of language a -in
"goes "the
thai it b"yorlJ mast.".y individual skilis and elements and requires to be
"r
properl| organised in varying social situations. The socio-linguistic Yt_"Y. of language
porit, ihut 6rr.ry utterance oJ"rrrc in a culturally determined but identifiable context of
situation in which individuals engage as participants to perform various roles'
What
factors'
they say and how they say it are determinedby the cultural as well as contextual
The psycholinguistic-sociolinguistic trend, which is also known as the
an opposing trend to the
integrative-ro.lolingniiic trend (McGirrell 1981), emerged as
psyJhometric-structluralist trend or the discrete - point/atomic approach to language
i.rtlng and acknowledge that language proficiency actualised in performance,-more than
is
the accumulation of dfscrete elements of language, and therefore language learning
testing therefore
more than gaining control on a set of habits or structures. Language
should be vlewed as a measure of communicative competence revealed by
a candidate's
performance in a given social situation (Howard 198^0,.Weir 1988)
in order for making the discussion "ury to follow, three separate aspects of the
the three
trend has been identified and discussed in separate headings as to bring about
(1) the
different though often complementary views together- The three aspects are
psycholinguisiic aspect, (2) the integralive aspect' and (3) the sociolinguistic aspect'

3.3.1 The psycholinguistic aspect

Influenced^by cognit"ive psychology and cognitive-lelTing theories,. psycholinguistics
'an
views mind as central ti leailing procisses. Unlike the behaviouristic approach
which regards language learning ui un external mechanistic process shaped by
environmental factJrs, psycholinluistic approach considers language. learning
as a
production,
;ilpi;. internal and abstract mental p.bb.tt. underlying the planning
percfption u"a of languig9. lhis aspect of language learning, which
iater'developed "o*piehension
as language acquisitioi and subseqt*lly second language acquisition,
J.u*, considerabty of Cn"o*skyan linguistics (Chomsky .1957) Ch.omskyan linguistics
-Saussnieun
i, iunau-.ntally different from linguistics in the sense that while the former

Vol.8, No. 1-2 December, 2003

Journal ofNELTA
 Language Testing: Then rnd Now/59

which exists in every normal human being, the later describes it as a physical process of
learning language structures and systems. Bloomfield (1933), who is regarded as the
mentor of structural linguistics, took a more concrete approach to language leaming and
emphasized the process of segmenting and classifring the physical features of
utterances (Chomsky's surface structures) and disregarded the underlying structures (or
what Chomsky calls deep structures) of language (Crystal 1997). Chomskyan theory
claims that every mind is pre-programmed with the ability of learning a language in
which internalisation of the underlying rules is the key to successful language learning
and adequate exposure to the target language the only condition for it.
Chomsky's work provided a basis for describing a child's (learner's) language as
a legitimate, rule governed, and consistently evolving system. This means that measures
of a child or learner's performance should have predictive ability of hislher competence,
which is unseen, unobservable, and abstract.
The psycholinguistic theory of language learning in 1960's led to the
development of various language-teaching schemes, which treated language learning as
a process of acquiring conscious control and understanding of language systems. The
theory advocatei that if a learner has a proper degree of cognitive control over the
language structures, facility will develop automatically with use of the language in
meaningful situations' (McGarrell 1 98 1 :29)
Language testing, under this thqory, shifted its emphasis from linguistic
accuracy to functional ability. Language tests reflected on problem-solving approaches
and were expected to reveal what underlying rules the learners had internalised. Target
language errbrs received a positive attitude and were considered to be an indication of
the learners' level of the transitional competence (Corder 1967) or 'interlanguage'
(Selinker 1972). Reliability and validity were still central to the language testing
(McGanell 1981)

3.3.2 The Integrative Aspect

The Integrative Aspect proposes to assess the candidate's capacity to use a number of
language abilities at the same time. In this respect, the integrative approach to testing
contrasts sharply with the discrete-point approach. The approach is commonly used in
proficiency tests because it reflects read language use more closely (Banerjee in Byram,
2OOO). Oller (1979) the mentor of what is popularly known as Unitary Competence
Hypothesis (UCH) proposed that there is a single, unitary factor that underlies language
pioficiency, and argued that argued that language competence is indivisible and that the
iour macro skills of listening, speaking reading and writing are so closely irser related
that it is appropriate to consider a general language proficiency factor rathpr than to
distinguish finely between them. As Oller (1979:112) writes
...the concept of an integrative test was born in contrast with the definition
of a discrete point test. If discrete items take languages skill apart,
integrative tests put it back iogether. Whereas discrete items attempt to test
knowledge of language one bit at a time, integrative tests attempt to assess a
-capacity
learner's to use many bits all at the same time, and possibly while
exercising ieveral presumed components of a grammatical system, and
perhaps more than bne of the traditionally recognised skills or aspects of
skills.

Journal ofNELTA Vol.8, No. 1-2 December, 2003

60 / Ram Ashish Giri

can test several

Oller's statement above suggests that the integrative apploach
traditionally named elements and skills or aspect because it
assumes that the general

language proficiency is a single principal facior underlying

all language elements and
of overall language
skills. He claims that the cloze or diciation can provide measures
as in discrete-point
proficiency, which, unlike testing one element or skill at a time
The cloze and
testing tradition, assesses several elements and/or skills simultaneously'
competence' but they are a
dictation, as he acknowledges may not test communicative
better guide to learners' uplitrrd. and potential communicative
ability' as they require
pressure, pragmatics' etc' They
them to perform under real life constraints such as time
mark and have respectable
are also practicable to administer, economical to set and
degree of reliability (Weir 1988).
a perceptual task, a
The cloze test has been criticised as a cognitive rather than
knowledge, unreliable, its
means to evaluate notional knowledge, rather than functional
validity in measuring EFL proficiency questionable, its validity
in measuring EFL
measure of comprehension
proficiency questiona[le, not a unitary technique and a crude
Morrow (1979), McNamara 1996)
lDavi"s t6g3, Weir 1988, Carroll 1980, Davies (1988) who
However, a serious criticism of oller's work comes from
evidence that there are two
says that the proposition of UCH conflicts with the
between knowing how to
competencies - reception and production. The difference
analyse input and knowing how to construct output
is stronger than the bluned
correspondence between the two processes which oller
claims to be the manifestation
are not a
of a single unitary factor. As for the cloze and dictation tests, they
because of their versatile
manifestation of ,rarious aspects of the ucH, rather, they,
nature, contain most language skills'

3.3.3 The sociolinguistic asPect

a language involves
The sociolinguistiJ aspect is based on the premise that knowing
more than the knowledge of language elements and skills.
It involves, alongside the
A candidate should'
knowledge of language code, ability to use the language in context'
and skills communicatively
therefore, be tested for his ability to use linguistic elements
and appropriately in a given situation. There are two aspects
of ability in using a given
language (a) knowledle of the language code (linguistic competence), and (b)
-which
knowledge of socio-linguistic factors includes knowledge of the culture and
social rules. A languagJ t.rt should measure ability to use language in both tlpes of
-test
knowledge becausJ a measuring the linguistic aspect of the ability only cannot
contexts'
account for how well the uSeI can communicate in given social
The sociolinguistic aspect draws upon Hymes's (1972) seminal work which
as its linguistic context'
suggested that the social context of a message is as important
and missing the socio-cultural clues of a message may result in
missing or
several projects of
misunderstanding.the message altogether. Hymes work inspired
and grammar
developing new teaching methods' language syllabuses, learning materials
Vol. 8, No. l-2 December, 2003
Journal ofNELTA
 Language Testing: Then and Now/61

books. Wilkins (1976), for example, outlined the notional categories of communicative
needs. Munby (1981) provided a scheme of elaborate specifications of the learners'
target language communicative needs, the Van Ek (i979) devised a communicative
language proficiency course for the Threshold Level, and Leech (1975) wrote an
elaborate communicative grammar. All this led to the development of teaching schemes
aiming at meeting the target language communicative needs of the learners.
So far as language testing is concerned, this aspect calls for testing schemes that
reveal the learners' ability in using the language in communicative situations. Language
tests should evaluate not only the learners' knowledge of the elements and skills but
also their ability to comprehend and produce utterances that are both situationally and
contextually appropriately.

4 The Communicative Language Testing

Several years after Hymes's two-dimensional model of communicative competence,
which consisted of both linguistic and sociolinguistic elements, several models- of
communicative language teaching and testing appeared in Europe as well as in North
America. In 1980s, particularly in early 1990s, language testing models such as those
proposed by Morrow (1979), Canale and Swain (1980) and Bachman (1990) advocated
that language testing should be concerned with both competence and performance, i.e.,
with (a) what the learner knows about the form and about how to use it appropriately,
and (b) the extent to which s/he actually demonstrates language knowledge in
meaningful performance (Weir, 1 990).
The communicative approach to language testing, though it is still said to fall in
the psycholinguistic-sociolinguistic trend, has following the works of Bachman (1990),
Bachman and Palmer (1996) and McNamara (1996), established itself as a promising
approach to second language testing. Bachman and Palmer have moved ahead of what
Hymes (1972) proposed and have added additional features in communicative
competence. They suggest that a communicative language test has four important
dimensions, namely (i) specificity of context, (ii) authenticity of materials, (iii)
authenticity of test tasks, and (iv) simulation of real life situation, and must
operationalise all four types of competencies - linguistic, sociolinguistic, discourse and
strategic. In other words, in a communicative language test, communicative
performance tasks should be representative of the types of tasks and materials the
candidates are likely to encounter in real-life situations and at the same time, corespond
to normal language use (Weir 1990).
Despite the progress in developing communicative language testing models,
only a small number of communicative language tests in second language have been
published. The UCLES's Certificates of Communicative Skills in English, and
Australian Educational Council's (AEC) ESL Scales and National language and
Literacy Institute of Australia's (NLLIA) ESL Development: Language and Literacy in
Schools are currently in use in the British and Australian education systems.

Journal ofNELTA Vol. I, No. I-2 December, 2003

62 / Ram Ashish Giri

5. I,anguage Testing TodaY

Today, lung.rug" testi particularly ESL/EFL tests, serve a wide variety of purposes
and
test is given for several formative or summative goals. The purposes' as the figure
below shows, can be summarised under general and specific purposes:

5.1 The Purpose of Assessment

General Specilic Description

General knowing learners' general ability in the language

Placement placing learners in ability groups
Administrative Certification providing proof of language ability
Promotion promoting learners to a higher level
Selection selecting able learners' for a purpose

Motivation encouraging learners to learn

Diagnosis ascertaining areas of difficulties in the language
Prognosis determining learners' readiness for a course
lnstructional evidence for progress checking whether learners are making progress
feedback to examinee monitoring learners' progress
Improvement evaluating teaching or a course

evaluation reviewing programme achievement

Research experimentation finding more about language and its learning
kntwledge knowing about communication and learning strategies

Figure: Language testing Purpose (based on Cohen 1994; and Madsen'1983)

These purposes are not always mutually exclusive to each other as one form
of assessment
may also be used for several other purposes. The assessment given for finding evidence
of
or
pfogress, for example, may also be used for researching students' learning styles
communication strategies. A test given for a general purpose may also serve the
purpose of
may also
motivating students towards learning. Similarly, a test given for testing achievement
be used for certification purposes.

WH- Paradigm of Assessment

5.2 The
The what, how, whe; and by whom of a test, as what is known as the 'WH- Paradigm
in
Assessment' (Cajkler and Addleman 1992) below suggests' largely depends on
why of
it.

Vol. 8, No. I-2 December, 2003

Journal ofNELTA
 Language Testing: Thcn and Now/63

WHAT
[skills or knowledge learned]

WHEN HOW fcriterion-referenced]

Icontinuous]
WHICH FORMAT WHO lteachersl
Iintegrative/subj ective]

WHY fformative] summative

administrators discrete-
point/objective

norm-referenced end-of-term
use of skills/application of knowledge

(Figure # based on Cajkler and Addleman 2000)

As the figure # above suggests, what is assessed (how well the learners are
achieving or how well they are learning), how is the results of the assessment is
interpreted (norm-referenced or criterion-referenced), which format/design of
assessment is used, and who conducts the assessment and who uses the results are all
dependent on why (purpose) of the assessment.
Of the three general and fourteen specific purposes of language assessment, as
outlined in the figure # above, a national test like the English test of the School Leaving
Certificate (SLC) Examination is given only for a general administrative purpose. Under
this general purpose, it serves two specific purposes - general proficiency and
certification or achievement. The general language proficiency is indicative of the
learners' target language linguistic competence as well as their ability to use this
competence in actual communicative situations. As a general proficiency test, the
English test of the SLC, is, as one would expect, a test of general language skills and
elements. The SLC reports an aggregate of scores only, and it does not contain
information on the outcomes of the test components nor does it provide descriptions of
abilities. The institutes, which may use the test results for selection purposes, make their
decisions on the basis of the aggregates.
Decision about certification is usually made on the basis of the achievement of the
target course. What type of certification (first, second or third division) is issued
depends on what has been achieved or learned from what has been taught in the course.
The English test of the SLC, in this sense, is an achievement test. In other words, the
SLC examination is a measure of achievement, the results of which are sometimes used
for screening or selection purposes. A score given on a subject is an indication of how
well a candidate has taken the examination. It is not usually an indication of how well
he has learned durins the course.

Journal ofNELTA Vol. B, No l-2 December, 2003

64 / Ram Ashish Ciri

Conclusion
'What to test' in language has always been an important consideration and is something
that has undergone reviews over time. The 20th century, particularly the second half
witnessed considerable changes and shifts in the theory and practice of language
teaching and testing. Though a timeline'for the changes cannot be explicitly drawn, it
appears that every decade of the second half of the 20th century experienced same sort
o?-change. In language testing 'What is to be tested' follows what goes in the teaching
and learning of that language. In other words, language testing tends to follow trends of
language teaching
The trends in the second or foreign language testing may be described as
analytical, global and communicative. The analytical approach to language testing
follows the description of language, which suggests that language consists of several
discrete systems and sub-systems, and language learning is gaining mastery over the
systems one by one. Testing, then, is employing objective tests, rvhich could be scored
consistently. The global approach advocates for combining various language sub-skills
and elements in testing because they are inter-related and interdependent, and for
integrating the test items into a total language event. The communicative approach, on
the other hand, views language as communication and language learning as developing
communicative competence, which is essential for enabling learners to use language in
the multiple functions it serves in the real life. In communicative testing, the best exams
are those that combine various sub-skills as we do when exchanging ideas orally or in
writing.
From a communicative perspective, language testing in ESL and EFL may
roughly be divided into pre-communicative and communicative stages. In the pre-
communicative language testing, the learners of a second/foreign language have to
demonstrate their practical command of skills acquired. This tradition continued
through structuralist, discrete-point framework of Lado (1961), pragmatic expectancy
g.u-iu, of Oller (Ig7g), integrative test structure of Carroll (1972) and
psycholinguistic performance model of Davies (1988). The communicative language
iestittg .utr b. linked to the seminal work of Hymes (1972), but a detailed testing
framework was only proposed in the much publicised work of Canale and Swain (i980)
which was later elaborated by Bachman (1990) and Bachman and Palmer (1996)'
Almost two decades ago, in order for a language test to be eclectic, a synthesis
of discrete-point approach and integrative procedures was viewed necessary. Twenty
years later on, when the psycholinguistic-sociolinguistic trend is still in vogue and the
tommunicative approactr to language testing is becoming increasingly popular, it is
assumed that no single language testing model or test 'can accommodate the wide
variety of test scenario'. The present research, therefore, is an attempt to justit that an
adaptation or combination of models or tests in different language areas and different
target situations should be made available in order to meet the local language testing
reouirements.

Vol. B, No. l-2 December, 2003

Journal ofNELTA
 Language Testing: Then and Now/65

In Nepal, for example, where ELT practitioners are still bound to the previous trend
and are not yet fully aware of the ways and means of communicative language
testing, a combination of all three aspects of the psycholinguistic-sociolinguistic
trend and the communicative approach to language testing can justifiably serve the
various purposes of the English test of the SLC.

References
Alderson J. Clapham, C. and Wall, D. (1995). Language Test Construction and Evaluation. Cambridge.
Cambridge University Press.
Alderson, J. C., Krahnke, K. J. and Stansfield, C. W. (19S7). Reviews of English Language Proficiency
Zesrs. Washington, DC. TESOL.
Allen, J. P. B. and Davies, A. (1977). Testing and Experimental Methods. Edinburgh Course in Applied
Linguistics. Vol. 4. London: Oxford University Press.
Asher, R.E. (1994xed.). The Encyclopaedia of Language and Linguistics. Oxford. Pergamon Press.
Bachman, L. F., Davidson, F., Ryan, K. and Choi, I. (1995). Studies in Language Testing. Vol. l-5.
Cambridge: Cambridge University Press.
Bachman, L. and Palmer, A. (1996). Language testing in practice: Designing and Developing Useful
Language Zesls. Oxford: Oxford University Press.
Bachman, L. F. (1990). Fundamental Considerations in Language Testing. Oxford: Oxford University
Press.
Banerjee, S. (2000). Integrated Testing. In Byram, M. (2000) Routledge Encyclopaedia of Language
Teaching and Learning. London. Routledge.
Bhadra, S. and Yadav, Y. (1988). Causes of Failures in the Proficiency Certi/icate Examinations.
Kathmandu: CERID.
Bloomfield, L. (1933). Language. New York. Holt, Rinehart and Winston
Brown, J. D., Cook, H. G., Lockhart, C. and Ranes, T. (1991). 'Soutleast Asian Languages Proficiency
Examinations ; ENC ED 365 | 52.
Brown, J. D. and Yamashita, S. O. (1995).'Language Testing in Japan.' ERICED 4001.13.
Cajkler, W. and Addleman, R. (2000). The Practice of Foreign Language Teaching. London. David
Fulton.
Canale, M. (1983). 'On Some Dimensions of Language Proficiency.'ln Oller, J. (ed.) (1983). Issues in
Language Testing Researcft. Rowley: Newbury House. 102-l15.
Canale, M. (1994). 'On Some Theoretical Framework for Language Proficiency.' In Rivera, C.
(ed.)(1994). Language Proficiency and Academic Achievemerl. Clevedon, Avon: Multilingual
Matters.
Canale, M. and Swain, M. (1980). Theoretical Bases for Communicative Approaches to Second
Language Teaching and Testing. Applied Linguisitcs. Nol. 1, No. l, Pp- l-47.
Carroll, B. J. and Hall, P. J. (1985). Make Your Own Language Tests. New York: Pergamon Press.
Carroll, J. B. (1983). 'Psychometric Theory and Language Testing.'ln Oller, J. (ed.) (19S3) Issues in
L anguage Testing Res earch. Ptowley: Newbury House. 19 5-207 .
CERID, (1989). Causes of Failure in English in the SLC Examination. Kathmandu: Tribhuvan
University.
Chomsky, N. (1957). Syntactic Structures. The Hague. Mouton
Clapham, C. & Wall, D. (eds.) (1996). Language Testing Update.. England: Centre for Research in
Language Education, Lancaster University.
Cohen. A. (1994). Testing Language Ability in the Classroom.Rowley: Newbury House Publishers, Inc.
Coombe, C. (1998). 'Current Trends in English Language Testing.' ERIC ED 428574.
Corder, S.P. (1973). Introducing Applied Linguistics. London. Penguin.
Crystal, D. (1997). The Cambridge Encyclopaedia of Language. Cambridge. Cambridge University Press.

Journal ofNELTA Vol.8, No. I-2 December, 2003

66 / Ram Ashish Giri

Ministry
Curriculum Development Centre (CDC). (lgg7). Secondary Education Curriculum. Kathmandu:
of Education, HMG/N.
Cziko, G. A. (1984). 'Improving the Psychometric, Criterion-referenced, and Practical Qualities in
Integrative Language Tests.' TESOL Quarterly. 16' 3:367-79'
Davey, L. dgg1). e Cise for a National Testing System. ENC Practical Assessment, Research and
Evaluation. ISSN 1096 0066.
Davey, L. and Monty, M. (1991). 'A Case Against a National Test.' ERIC Practical Assessment, Research
and Evaluatior, ISSN 1096 0066.
Davies, A.(1963) (ed.). Language Testing Symposium. A Psycholinguistic Approach. Oxford' Oxford
University Press.
Davies, A. (1983).'Operationalising Uncertainty in Language Testing: An Argument in Favour of
Content Validity.' Language Testing. 5. 1:32-48.
Davies, A. ( 1990). Principles of Language Testing. oxford. Black Basil.
EzuC. (1995). 'The Cost of aNational Examination.' ERIC/AE Ed' 385611'
Farhady, H. (1979).'The Disjunctive Fallacy between Discrete-point and Integrative Tests.'
TESOL

Quarterly. 16. 1:43-59

Farhady, H. (1983). 'New Directions for ESL Proficiency Testing.' In Oller, J. (ed.)
(1983). Issues in
Language Testing Research. Rowley: Newbury House' 34-46'
,A Text Driven Method for the Deletion Procedure in Cloze Passages.' ERIC EJ
Farhady, H. if qgO).
529503.
Fries, C. (1945). Teaching and Learning of English as a Foreign Language' New York' Harcourt
Brace'
f'ries, C. (1952). The Stn.rcture of English. An Introduction to the Construction of English Sentences.
New York. Harcourt Brace.
B. (1988). Ihriting English Language Tesls. London: ELBS'
Heaton, J.
Howard, F. (1980). testing-Comiunicative Ploficiency in French as a Second Language: A Search
for
Procedures. canadian Modern Language Review vol 36, No. 2, Pp. 272-280.
Hughes, A. (1986). Testingfor Language Teachers. Cambridge: Cambridge University Press..
Hyiies, D. (1972). on Communlcative Competence. In Pride and Holmes (eds.) Sociolinguistics.
London. Penguin.
Ingram, D.E. (1934). Using Proficiency Rating Scales with High School Foreign Language Learners'
ERIC ED 249761,
Ingram, D. E. (1985). 'Assessing Proficiency: An Overview of Some Aspects of Testing.'In Hyltenstam,
K. and pienemann, V. (eOs) (1985). Modelling and Assessing Second Language Acquisition.
Clevedon Avon: Multilingual Matters.
Jones, R. L. (1980). 'An Intemational.Survey of Research in Language Testing'' ERIC Ed. 192579
Lado, R. (1961). Language Testing. London. Longman
Lee, S. (1996). 'The Co-ncunent Validity of the Cloze Test with Essay Test Among Korean Students.'
ERIC Ed. 4167 13.
Leech, G. (1975). The communicative Grammar of English. London. Longman
Leedy, p. D. (1981). How to Read Research and (Jnderstand il. New York: MacMillan Publishing
Co.
Inc.
Madsen, H.s. (19s3). Techniques in Testing. oxford. oxford University Press
McGarrell, H. (1931). Language Testing: A Historical Perspective. In Medium Vol' 6, No' 4
McNamara, T. (2000/. Language Testing. Oxford: Oxford University Press.
McNamara, T. F. (1996). Measuring Second Language Performance.London: Longman'
Morrow, K. (1979). Communicative Language Testing: Revolution or evolution? In Brumfit, C' and
Johnson, K. (eds.). The Communicative Approach to Language Teaching. Oxford. Oxford
University Press.
Munby, J. (1981). Communicative Syllabus Design. A Sociolinguistic Model for Defining the Content of
- purpose-Specific
Language Programmes. London. Cambridge University Press.

Vol. 8, No. l-2 December, 2003

Journal ofNELTA
 Language Testing: Then and Now/67

Oller, J. (1979). 'The Psychology of Language and Contrastive Linguistics: The Research and the
Debate'. ENC EJ206643.
Oller, J. (1983). Issues in Language Testing Research. Rowley: Newbury House.
Oller, J. (1919). Language Tests at School: A Pragmatic Apprctach. England: Longman.
Oller, J. ( I 99 I ). 'Current Research/Development in Language Testing.' ERIC ED 365146.
Oller, J. (1992). 'Foreign Language Testing; Part 2: Its Depth. ADFL Bulletin,22:5-13
Rea-Deakens, P. and Germaine, K. (1996). Evaluation. Oxford: Oxford Universify Press.
Selinker, L. (1972) Interlanguage. IRAL Vol 10. Pp-209-231
Skinner, B. F. (1957) Verbal Behaviour. New York: Appleton, Century, Crofts.
Spolsky, B. (1969).'Introduction: Linguists and Language Testers.'In Spolsky, B. (ed.) (1969).
Approaches to language Testing. IAashington, D. C. Center of Applied Linguistics. l-19.
Spolsky, B. (1975). Language Testing: Art or Science? Key Note Address at Association Internationale
de Linguistique Appliquece World Congress, France.
Spolsky, B. (1978)(ed.). Approached to Language Testing. Papers in Applied Linguistics. ERIC ED
16548
Spolsky, B. (1993 Testing of English of Foreign Students in 1930. EzuC 2002.
Spols\y, B. (i995). Measured Words: The Development of Objective Language Testing. Oxford. Oxford
University Press.
Van Ek, J. A. (1979). The Threshold Level. Council ofEurope. Transbourg, 1979
Weir, C.J.( 1988). The Specification, realisation, and validation of an English language proficiency test.
In Hughes, A. (ed.) Testing English for University study. ELT Document 127. London. Modern
English Publications/The British Council.
Weir, G. (1990). Communicative Language Testing. New York; Prentice Hall.
White, E.M. (1994). Teaching and Assessing Writing. San Francisco: Jossey - Bass Publishers.
Wilde, J. and Sockey, S. (1995). Evaluation Handbook. Albuquerque: Evaluation Assistance Centre, New
Mexico High Land University.
Wilkins, D. (1976). Notional Syllabus. A Taxonomy and its relevance to Foreign Language Curiculum
Development. London. Oxford University Press.

Contact: Ram.Giri@research.vu.edu.au
(Currently Mr. Giri is a Doctoral student at Victoria Universtiy, Australia)

Journal ofNELTA Vol. B, No. l-2 December. 2003

View publication stats