Language Testing:
TESTING ORAL ABILITY
The objective of teaching spoken Language is the development of the ability to
interact successfully in that language .This involves comprehension as well as
production .
The basic problem in testing oral ability:
 We want to set representative tasks of the population of oral tasks that we
expect candidates to be able to perform.
 The tasks should elicit behavior which truly represents the candidates’ ability.
 The samples of behavior can and will be scored validity and reliably
1. Representative tasks
• We will begin by looking at the specified content of the Cambridge CCSE Test of
Oral Interaction
1.Operations
- Expressing - Directing
- Describing - Eliciting
- Narration
- Reporting
2. Types of text : Discussion
3. Addressees : ‘Interlocutor’ (teacher from candidate’s school) and one
fellow candidate
4. Topics :
Unspecified
5. Dialect, Accent and Style : Also unspecified
It can be seen that the content specifications are similar to those for the Test of
Writing
The categorization of the operation (skills) is based on Bygate (1987)
 Skills
 Informational skills: providing personal and non-personal information,
describing, giving instructions, comparing, eliciting , apologizing, etc…
 Interactional skills : expressing agreement, disagreement , indicating uncertainty
or understanding …
• Skills in managing interactions: initiating interactions, changing the topic of
interaction, taking turn in interaction , …
 Type of text
Presentation (monologue)
Discussion
Conversation
Service encounter
Interview
 Other speakers (addressees): Maybe equal or higher status
Maybe known or unknown
 Topics: Topics which are familiar and interesting to the candidates
 Dialect : Standard British English or Standard American English

Accent : Standard American


Style: Formal or informal
Vocabulary range :Non-technical except as the result of preparation for a

presentation
Rate of speech: will vary according to task
2. Elicit a valid sample of oral ability
A) Choose appropriate techniques
Three general formats are presented here: interview; interaction
with fellow candidates; responses to audio-or video-recorded stimuli
 Format 1 – interview
useful techniques are:
1. Questions and requests for information
Yes/No questions should generally be avoided
2. Requests of elaboration: “What exactly do you mean?”
3.
Appearing not to understand :
in order to see if the candidates can cope with being misunderstood
4. Invitation to ask questions :“Is there anything you’d like to ask me?”
5. Interruption: to see how the candidate deals with this
6. Abrupt change the topic: to see how the candidate deals with this
7. Pictures: single picture are particularly useful for eliciting description
8. Role play:
this allows the ready elicitation of other language functions
9. Interpreting: you can see the example how to conduct this on page 120
10.Prepared monologue
11.Reading aloud
 Format 2 – Interaction with fellow candidates
Possible techniques:
1.Discussion: Tasks may require the candidates to go beyond discussion and, for
example, take a decision
2.Role play: Role play can be carried out by two candidates with the testers as an
observer
 Format 3 – Responses to audio-or video-recording
 Described situations
 Remarks in isolation to respond to
 Simulated conversation
see page:122 for examples
B) Plan and structure the testing carefully (p.124)
 Make the oral test as long as is feasible
 Plan the test carefully
 Give the candidate as many ‘fresh starts’ as possible
 Use the second tester for interviews
 Set only tasks and topics that would be expected to cause candidates no
difficulty in their own language
 Carry out the interview in quiet room with good acoustics
 Put candidates at their ease so that they can show what they are capable of
 Collect enough relevant information
 Do not talk too much
 Select interviewers carefully and train them
3. Ensure valid and reliable scoring
Create appropriate scales of scoring
Rating scales may be holistic or analytic. Skills required from ss to show are:
ACCURACY: Pronunciation must be clearly intelligible even if some influences from
L1 remain
SIZE: Must be capable of making lengthy and complex contribution where
appropriate
APPROPRIACY: The use of language must be generally appropriate to function and
to context
RANGE:
A wide range of language must be available to the candidate
FLEXIBILITY: There must be consistent evidence of the ability to ‘turn-take’ in a
conversation and to adapt to new topics or changes of direction
Creating rating scale:
1. Calibrate the scale to be used
Generally the same procedures are followed in calibrating speaking scales as
were described for writing scales
2. Train scorers (as opposed to interviewers)
The training of interviewers has already been outlined. Where raters are
involved in the rating of responses to audio-or video-recorded stimuli
3. Follow acceptable scoring procedures
Great care must be taken to ignore personal qualities of the candidates that
irrelevant to an assessment of their language ability
Conclusion
The accurate measurement of oral ability is not easy. It takes
considerable time and effort, including training, to obtain valid and
reliable results. Nevertheless, where a test is high stakes, or
backwash is an important consideration, the investment of such
time and effort may be considered necessary. We are reminded
that the appropriateness of content, of rating scale levels, and of
elicitation techniques used in oral testing will depend upon the
needs of individual institutions or organizations.