You are on page 1of 8

TESTING LANGUAGE ABILITY

We have already seen the need for systematic evaluation of all aspects that influence learning so that
we could explain why certain things are going well and certain things are not. Now we will discuss the
evaluation of a very important component of the teaching-learning process, i.e. the learning outcome. In
fact, the whole exercise of education would be futile if it were not for some desirable learning.
Therefore in a sense the result of education, in this case, the learner's language ability, is the most
important.

Importance of testing

Knowing how well a learner has performed or learned in a programme of study is an integral part of
education. In fact, without a proper knowledge of where the learner was, how far s/he has got, how far
s/he can go. s/he won't be encouraged to continue. Therefore, testing/appraisal or stock-taking is
inextricably linked with teaching.

Usually, as it happens in the Indian context, tests/exams are extremely important and students' lives and
future careers are dependent on test performance. Therefore, as it rightly happens, teaching is done for
exams. But the question that comes up is, what do exams test? Most often it is memory. They test the
recall of content: what is read and understood, e.g. in the case of language learning, essays, poems,
plays, etc., not how well the learner can use the language in different situations for different purposes,
i..e the skills of reading, writing, listening, speaking, etc. In typical exams, testing of language skills is
minimal. What happens in class, therefore, is obvious. The teacher and students together prepare for
these exams. They learn to reproduce answers to expected questions. These ready-made answers come
from the teacher's own notes, bazaar guidebooks, etc. Whatever exposure to skill-use/practice students
get during this exam-preparation, which happens throughout the year, seems to be accidental. In fact, it
is the teacher who is getting a lot of practice in speaking, reading, etc! Students are just passive
receivers of pre-digested content.

The situation that obtains, then, is one where exam requirements totally decide what happens in class.
This influence or effect which exams have on teaching is called the backwash or washback of exams.
We are not of course talking about removing that effect. Exams are a very important component of
education and therefore they have a powerful backwash on the education system. The question to
consider is how beneficial or harmful the backwash is, as it exists.

We should remember one point in this context. Changing an exam is possibly the most powerful means
of bringing about improvements in teaching-learning in class. By the same token, if materials are
changed to reflect a new approach to the syllabus, and teachers are trained in the new materials and
methodology, and no corresponding changes are made in the exam, all this enterprise will be a waste.

Different types of Tests

We have so far discussed the importance of exams in our system and the kind of backwash it can have
on teaching-learning in the classroom. There are other purposes for which language testing is carried
out. Different types of tests are administered for different purposes. As prospective language testers,
there is a need for you to know what function each type of test serves and at what stages of the course
they should be administered for gathering information about students.

Summative/final exams are administered at the end of a course to find out student achievement i.e.
how much and how well students have learnt to use the language. While these exams are effective for
certification purposes and to decide whether students can move on to a higher level, a lot of
information that a summative test can provide is often wasted, because it does not feed back into the
learning process. Also, as we all know, these tests put a lot of stress on teachers and students and may
not tap the full potential of students.

Progress Test is another kind of achievement test which aims to find out, on an on-going basis,
information about how well classes as a whole and individual students have worked. These, in our
school/college contexts are the unit tests, half-yearly/annual exams which are class-based and which
teachers conduct during the course. These can provide a great deal of information about the objectives
(how realistic they are), materials used (appropriacy, interest, challenge), methodology/strategies used
(techniques, class-organisation, interaction) and student outcomes (sub-skills, writing/reading
strategies) not only to the teacher but also to the student. When linked with self-assessment, feedback
can help learners to identify their own problems and to set their own goals for the future.

Proficiency tests aim to describe what students are capable of doing regardless of any training they
might have had in the language. They enable students to have some proof of their ability which other
Boards/Universities/ potential employers can use for their own purposes. Although the tests are not
based on the content of a particular course, they may exercise considerable influence over the method
and content of language courses in general and on courses that are taught for such professional tests.
For example, coaching classes for training learners in taking, say, the Test of English as a Foreign
Language (TOEFL) or the International English Language Testing System (IELTS), may have a beneficial or
a harmful backwash effect, depending on the content and format of the test.

Entry/placement tests will indicate at which level a learner will learn most effectively in a situation
where there are different levels and streams. The aim is to produce groups which are homogeneous so
that teacher time can be used most efficiently. However, the need for these tests can be questioned on
the grounds that students learn best where there are learners of different abilities in a class, which can
be exploited through cooperative/collaborative learning.

Diagnostic tests, as the name suggests, are used to find out students' problem areas as well as areas of
strength. Diagnostic information is vital for teachers in order to design further remedial activities. This
information can also be useful for learners as they can analyse their own strengths and weaknesses.
The role of the teacher in testing

We have said in the earlier section that tests, i.e. final exams, which are conducted by Boards and
Universities, determine to a very large extent what teachers and learners do in class. If the exam is a
good one, that is to say a language skill-based one where learners are required to read, write, speak,
listen at the time of the exam, it will make students/teachers engage in similar activities in class.
Teachers will also construct similar class tests. Then there is a lot that the teacher needs to know and
can do about such language tests. If on the other hand, the exams are "traditional' memory-based ones,
what can teachers do? It would be naive to argue that they should prepare students in class for anything
other than those things on which they will be tested. But in spite of this situation teachers can do their
bit to change the final exams. They can build a pressure group, keep a dialogue going with the Board or
syllabus committees and so on. In any case teachers will have to know what they can do in class by way
of better testing (when the time does come) and in fact know what changes to press for. The point we
are trying to make here is that however externally oriented the exam system is, the teacher has a
definite and a very important role to play in initiating change and in creating a more conducive language
classroom for teaching and testing.

It is necessary to carry out this business of testing in a much more informed and professional way than it
has been hitherto done. Further, in the context of a learner-centred classroom, teachers have a
responsibility towards the learner in that we have to involve the learner too in decision making about
what to test and how to test, as well as in the interpretation of test results.

Progress testing in the classroom

Summative evaluation is an end-of-course test that is usually controlled by a Board or a University.


Teachers have little or no say in the design of such tests and in the way results are used. It is usually to
assign students to a pass, fail, or distinction category.

However, teachers are responsible for tests that they conduct in class during the course. Very often
this classroom-based testing/evaluation also includes (informal) observation, assignments, class
records, projects, etc. for assessing learner progress or for making decisions about whether a learner
could move on to another class. But these internal measures are treated very casually and have never
gained the respectability of the external public exams. In fact, even these internal evaluations have
tended to be product oriented and reflect the design and criteria of assessment of the 'official' final
exam almost totally. In other words, teachers and students prepare for the final exam which may be
months away by rehearsing the exam in their class several times over during unit tests and termly
exams.

This anxiety to face the final exam (a monster, as it were) results in two major problems with regard to
class tests. One, the tests during the course are modelled on the final exam in terms of not only the
content (including the skills tested and their relative weight) but also of the level. This is clearly
questionable. Tasks /questions are pitched at the level of the final exam and do not reflect the stage of
language development which the course has envisaged, at a given time.
Another associated problem is that of over-emphasising those skills that are covered in the final exam,
to the neglect of other skills. One national phenomenon we experience is of oral skills not being given
any attention even during the course, because they are not important from the exam point of view.

The syllabus usually advocates the development of all the four skills. But the final exam is a paper-pencil
content-based test. What is the washback of this on teaching-learning in class?

The activities in class centre round the exam demands and therefore no systematic and deliberate
exposure to and training in the different skills are made available. Note the words 'deliberate' and
"systematic'. Students do now and then get to read, listen and speak depending on the ingenuity and
interest of the teacher, the type of text being dealt with, the size of the class, etc. This is not planned for.
Therefore monitoring and evaluation of how these skills are being developed seem to lie ‘outside the
syllabus'.

The other serious problem with classroom testing being over-burdened with the concerns of the final
exam is with regard to the use to which the test scores/results are put. Scores on class tests/exams are
treated the same way as those on final exams. Therefore very often discussions about why a particular
score was given and how improvement is possible are not seen as an important part of the course.

This essentially necessitates giving feedback, qualitative comments whenever possible, not just marks,
which will make sense to the students. This feedback should be negotiated between the teacher and
students. We are, therefore, talking about formative testing in the classroom which is of much more
value so that students can progress effectively during the course.

We can summarise the three points that seem to be crucial for classroom testing, which is also known as
Progress Testing.

1. Include in class tests all skills. It is necessary to train students in all the skills in an integrated
way, even if the final exam may not include it for practical, administrative reasons,

2. Give feedback on student performance. Maintain a record of progress, to see how much each
individual progresses during the course. Make the evaluation transparent to all those who are involved
in it.

3. Use a variety of ways to evaluate progress i.e. diaries, self-evaluation checklists, classroom
observation, questionnaires, tests and others. A sample of these is available in Block V of Materials for
the Teaching of English.

Tips for testing different skills

1. Test only the skill(s) one wants to test. This may sound commonplace but it is not very difficult to find
test items that claim to test a particular skill but in actuality test something else e.g testing reading
through seen passages (which tests prior knowledge/memory), answering long-answer questions on
unseen texts (which tests writing and reading), testing writing through a task which demands reading a
longish text as the basis for writing (which tests reading and writing), the ability to argue for and against,
in a pre-prepared talk/speech (which tests memory, skills of presentation, etc.).

2. Avoid excessive overlaps in structures, skills or task being assessed in different parts of the test. For
example, "argumentation' may appear in the reading text, writing tasks, debates, etc. Letters' may be
overdone as a format in different sections of a test. "Reported Speech' may be the feature of many tasks
under different skill areas.

3. Make sure the rubrics (instructions) are in a language that is easily accessible to all the students. If you
have any doubt about some section of the student group not getting at the task, you may have to
reword/simplify it. Tests for ESL students could also have rubrics in their mother tongue or a common
regional language, provided that they are clear and concise!

4. If it is a task in oral-interaction involving listening and speaking, the criteria will need to be derived
taking into account the participants' ability to interact, negotiate, take turns, etc. Otherwise one student
may be penalised because of another student's inability or even "bad behaviour'!

5. When the focus of a task is on a skill, look at the "appropriacy' of grammar and vocabulary in the
given context for the specified purpose. i.e. how appropriately a particular structure or word is used
rather than the accuracy of the construction. Accuracy of different elements such as spelling,
pronunciation, stress, the use of articles, prepositions, etc. can be tested separately in other tests which
are very useful for diagnostic purposes/for identifying students' specific problem areas.

6. The need to bias for test (i.e. to get the best out of our students), implies that we should give the
student adequate chance to perform so that our judgement of his/her language ability is accurate.
Therefore,we have to take enough samples of a student's work at different stages of the course. Since
we have a whole year in which to get a complete and accurate picture of the student, we should stagger
the tests and ensure that information about different aspects of their language ability on different types
of tasks over a period of time is available.

7. Informal assessment is done through observation, looking at student assignments/projects,


journals/diaries, etc. However it will need to be carried out systematically. On the other hand formal
assessment, where we elicit students' language behaviour in response to a stimulus/task (at the time of
testing), is carried out for the express purpose of assessment. Since both these types of assessment have
their own advantages and disadvantages, it is useful to think about what things we are going to assess
consciously and formally, and which things we are going to get an impression of, informally, as well as
when during the course.

8. Lastly, it is very difficult to write good test items. Therefore, let us not be discouraged if after a
sound theoretical background, our tasks turn out to be poor. Writing good tests items often comes with
experience, of course, backed by adequate "testing' knowledge. It is also best done in groups, since
teachers with similar concerns can come up with a number of task types and the larger the sample of
test types, the better the chance for the student to perform well. Also there is a need for a critical
perspective both in writing test tasks and in
scrutinising those prepared by yourself and others. There is a greater need however for teachers to
think of "authentic' tasks that work in class, in terms of students' linguistic level, motivation, interest etc.
More than anything, teachers know which type(s) of task students enjoy doing. Make sure that testing is
fun especially in classroom testing.

Review question I

Which are the type(s) of test that are most relevant to you as a would-be teacher? Why?

Achievement tests, both summative and progress tests, are relevant in a situation where a course is
offered. We need to prepare students for the final/summative test, and therefore we need to keep track
of how they progress towards the target objectives. By the same token, progress tests which give us
information about on-the-way achievement or the means to the end are equally important, so that both
the teacher and students can apply corrective measures during the course to achieve the target. Also
diagnostic tests, which give us focused feedback on what students are good at and what they are not
good at, are useful since these enable us to think of alternative ways of teaching and learning for
achieving the objectives.

Review question II

Many teachers complain that testing wastes time that can be usefully spent on teaching. What is your
view on this?

Progress tests, as we have learnt, are a central part of teaching and learning. There is in fact a very thin
line between teaching and classroom testing. We need to use information on students' performance in
tasks, during teaching as well as testing, to help them improve their language learning.

General considerations in testing

(This section is for your general reference only, not strictly for exams)

[The following are some general considerations in testing, that will give us guidelines for better testing
procedures.

1. Bias for best

Tests usually create negative reactions because they try to trap the student or worse still, reinforce what
students don't know. It is very easy for us to set tasks which are beyond the abilities of students, but
which give us an upper hand because we can assure ourselves that we know better than them! We
should always bias for best i.e. give students a chance to demonstrate their best or their optimal ability.
This, of course, goes against the popular practice of using tests as a weapon or for wielding power.

2. Test those abilities that you want to develop in class

We talked of the powerful influence or backwash tests have on what happens in the classroom. We
should exploit that influence in a desirable direction. For example, if you want to encourage students'
own writing, then test that ability and not the ability to reproduce memorised answers. If you want
students to develop speaking abilities then test just that. This argument may seem commonplace but it
is surprising how often it is not followed. Most of our syllabuses clearly state as the objectives of the
course the ability to read, write, speak, etc. And yet these abilities are not tested.

3. Make tests as direct as possible

There is a strong argument for making tests as direct as possible. Direct tests involve getting students to
demonstrate precisely those skills which we wish to measure. If we would like to know how well
students can speak, we get them to speak. If we want to know how well they can scan through a text to
get at specific information, then we get them to actually read through a text and do this task (at the
time of the exam, not before). The tasks that we use in the test would therefore have to be as authentic
or real-life as possible. Indirect tests attempt to measure abilities that underlie the skills in which we are
interested. For example, a multiple choice item to test the writing ability (such as Which of these best
completes the sentence?), or a stress-marking task in a paper-pencil test, are indirect tests. Although
these enabling abilities are related to the skills in question, it is not the same thing as testing the main
skill. Especially when it comes to the interpretation of scores on indirect tests, we are stumped because
we are not very sure whether a high score on an indirect test means good performance on the skill.

4. Make testing criterion-referenced

This consideration actually does not relate to the actual test or its content but to the way scores on tests
are interpreted. The yardstick we use to make judgements about student performance can be seen to
fall into two main categories : if we compare one student with another or with that mythical creature
called the average student, we are judging against a norm. This kind of testing is norm-referenced. On
the other hand if we are clear about what students should be able to do and with what degree of
success at the end of a course, and judge their performance against that criterion, the testing is
criterion-referenced.

Another concept that has gained currency in the context of learner-centred teaching-learning is
Ipsative-referenced assessment by which a learner's performance is measured against his/her own
previous performance (Ipsative derives from the Latin word for 'self')- It gauges individual progress over
time. Such a process, which accommodates the unique, idiosyncratic behaviour of each learner, is clearly
learner-centred and captures the gains made, however small they may be.

All these point to the need for clearcut specifications before constructing test tasks, so that we know for
certain what we are testing, for absence of such objectives will cloud the interpretation of test scores.

5. Grade tests wherever possible

One observation which most teachers agree on is that their class is heterogeneous, in terms of language
competence. Therefore (conscientious) teachers are concerned about tests not providing enough
challenge for the better students but at the same time being unfair to the weak ones. A test trying to
accommodate the so-called "average learner' cannot satisfy the others and probably does not satisfy
any one! A way out of this would be to think of levels built into a single test which would be transparent
to the teacher and the learner. Learners could be encouraged and trained to enter the level they think is
appropriate to them. This prevents learners from having to take a test which is far below their
capabilities or which encourages them to blindly give inappropriate answers or which demotivates them
because they cannot attempt even half the number of questions.

An alternative to questions at different levels is to select a set of tasks across a range of ability and make
a distinction in the quality of performance. Here all students would take the same test but different
levels of performance would be expected at different levels.

6. Ensure that the testing method/format is familiar to students

For the test to do its job well, i.e. to measure as accurately as possible the ability in question, the
students taking the test will need to be familiar with the method of testing. The method includes, for
example, the format (multiple choice, short answer, long answers, true/false etc.), the rubrics (the
instructions for doing the task, example or sample item) and how the performance is assessed (the
criteria for a high and a low score, the assessment scale : qualitative and quantitative). Unless students
know these things before they take the test, the score may not reflect the "true' ability being tested but
their ability to understand instructions or to figure out how a question has to be answered! And what is
more, we will never know what the score is for. Therefore, we will have to ensure that students (and of
course teachers) know what the test demands of them. This is particularly important when a new testing
system is introduced which is different from the old one.]

Sources (of both Evaluation and Testing)

Alderson, J.C. (ed). 1985. Evaluation : Lancaster Practical Papers in English Language Education 6.
Oxford : Pergamon.

Harris, M & P. McCann. 1994. Assessment, London : Heinemann.

Hughes, A. 1989. Testing for Language Teachers. Cambridge : CUP

Nunan, D & C. Lamb., 1996. The Self-directed Teacher : Managing the Learning Process. CUP.

Rea-Dickins, P. &. K. Germaine. 1-992. Evaluation. Oxford : OUP.

Sutton, R. 1991. Assessment : A Framework for Teachers. London : Routledge.

Weir, C.J. 1993. Understanding and Developing Language Tests. London : Prentice Hall.

You might also like