Applied Linguistics 2

(2021-2022)

7. Language Testing
(Testing Techniques)
Dr. Nada Alqarni
Here are some well-known testing techniques:
 Gap-filling (isolated sentences)
• Example: Fill in the gap with the right word.
• If it rains, you can get _____.

 Matching of words or sentences (usually synonyms or opposite meanings)

• Example: Match each word from the first list with a word with the same meaning from the second list
• large, unhappy, a lot of, little
• small, many, big, sad

 Multiple-choice
• Example: Fill in the blank with a suitable word from the list below.
• (1) hot (2) wet (3) bored (4) frozen
• If it rains, you can get _____.
Cloze (a passage)

• Example: Fill in the gaps with suitable words.

• A cold front is approaching from _____ west and we can therefore expect lower _____
tomorrow throughout the country. There is _____ to be some rain in the morning,
_____ it will become brighter later in the _____.

Syntactic transformations

• Example: Put the sentence into the passive

• The dog bit the man.

Dictation

• The tester dictates a passage and the test takers write it down. Usually the passage is read
through once first at normal speed before candidates write, is repeated slowly in sections with
repetition while they write, and is finally read through again at normal speed. (For specific
focus on spelling, a set of isolated words can be dictated.)

Listening or reading comprehension questions

• Example: Was the door open or closed when Jim arrived at the house? (Yes/No)

• The jacket was lying under the chair. True or false? (True/False)

• Where was the jacket when Jim entered the dining room? (Factual question)

• Why do you think Jim shuddered when he saw the jacket? (Opinion question)
Guided writing (sentences or passages)

• Example: (Based on a written text) Rewrite the passage from the perspective of the
woman referred to in line 3, making any necessary changes.

Translation

• The candidate is required to translate words, expressions, sentences or a passage into

or out of the foreign language. Sometimes dictionaries may be provided, sometimes
not.
Task-based writing

• A simulated real-life piece of writing, often a letter or report, is to be

produced with a specified goal, addressee and context.

Essay writing

• Example: Write an essay of approximately 2,000 words on EITHER

‘The role of the monarchy in Britain today’ OR ‘The role of drugs in
sport today’.
• Gap-filling, matching and multiple-choice are classic discrete-point test techniques.

• Cloze is claimed to combine elements of discrete-point and integrative testing because

although it consists of separate items, these items are not completely independent of one
another and an understanding of the passage as a whole is required.

• Dictation of a passage is an integrated test but dictation of single words is a discrete-point test.

• Listening and reading comprehension tests often combine discrete-point techniques (yes/no,
true/false and factual questions) with integrative techniques (evaluative or opinion questions).

• Guided writing, translation, task-based writing and essay writing are integrated tests.

• All the tests listed above are written tests and for this reason are better guides to written
proficiency than to oral proficiency, which is usually tested directly.
 Tests Which Focus on Receptive Skills
• Sometimes one may wish to test mainly listening or reading comprehension, respectively,
while minimising the use of speaking or writing.

• This is often done by setting yes/no or true/false questions or factual questions requiring a
short answer based on a recorded spoken text or a reading passage, respectively.

• The longer the answers the questions require, the more the test becomes an integrated test
of both receptive and productive ability (which may well be desired).

• Evaluative (opinion) questions usually require longer answers than factual questions.
Multiple-choice questions based on a reading passage or a recorded spoken test also test
comprehension without demands on productive ability.

Multiple-choice tests are quick and easy to mark and they are objective, but these
advantages are partly offset by the difficulty of constructing the test.

The great problem is to produce options which to the learner should all seem equally
plausible. This is in fact very difficult to achieve, particularly if the test writer has no
assistance.
When designing multiple-choice tests there are therefore a number of principles to abide by:

• There should be only one correct option per item.

• There should be no fewer than four items to choose from (otherwise it may be too easy to guess)
and no more than five (because in practice it is often difficult to find more than five plausible
options).

• The sentence stem should provide the full context necessary to solve the task and conversely the
options should be dependent on the context created by the stem.

• All the options should be formally comparable so that test takers cannot make shrewd guesses at
the odd one out.

• The stems and the options should also be kept reasonably short. (The form of the stem can be
varied, however. It may be an incomplete statement, a question, a gapped sentence or a sentence in
which an element is to be replaced by one of the options.)
 Test Selectivity
All language testing is necessarily selective along a number of dimensions, especially the following:

 ‘language selectivity’

• This refers to the range of grammatical structures, vocabulary, pragmatic functions and genres as well as the
number of language varieties, styles and registers which are tested.

 ‘mode selectivity’

• This refers to whether the test is a written or a spoken test.

 ‘skill selectivity’

• This refers to which of the four skills: listening, speaking, reading and writing, are tested.

 ‘test design selectivity’

• This refers to the tasks the test taker has to perform, for example a gap-filling exercise, translation or a dictation.
‘topic selectivity’

• This refers to what the thematic content of the test is, for example an essay on
‘Drugs in Sport’ versus an essay on ‘The Role of the Monarchy in Britain Today’.

‘test-conditions selectivity’

• This refers, for example, to whether the test is an open-book exam or not, whether
it is done under time pressure or not and, for an oral presentation, whether
preparation time is allowed or not.
• When a test is highly selective on a number of these dimensions, great care must be
exercised in generalising its results to other contexts.

• Thus, learners who score well on a grammar test which tests only prepositions might
not have done so well if the test had been on the passive and vice versa.

• A written test is not necessarily a good guide to spoken ability and vice versa.

• A test based on receptive skills (listening and reading) may not always be a good
guide to proficiency in the productive skills (speaking and writing) and vice versa.
• Transferability of proficiency from one task to another should not be assumed either. For
instance, good translators are not necessarily good letter writers and good
conversationalists are not necessarily good negotiators or good lecturers.

• Discrete-point tests may not always be the best of guides to integrated language use, and
knowledge about the language may not be transferable to performance in the language.

• Tests done under time pressure may not be a good guide to using the language without time
pressure and vice versa. Oral ability tested in a one-to-one interview may be a poor guide
to conversing against background noise or chairing a meeting.