Summarization of H.

Douglas Brown's
Language Assessment: Principle and Classroom Practice

Assessing Speaking
 Members
Amonchet Puwadethanyakarn
633080263-9
Thanaporn Nambutr 633080235-4

Nichakorn Dumklin 633080232-0

Suphaprat Vudhavanna 633080602-3
 Challenges
From a pragmatic view of language performance,
listening and speaking are almost always closely
interrelated.
It is challenging to separate oral task that is
included with listening comprehension.

The design of elicitation techniques

Our design must elicit and specify the content
that we want to assess.

More open-ended, more challenging in scoring

due to the testee's freedom of choice
Basic Types of
Speaking

Initiative Intensive Responsive Interactive Extensive

Ability to imitate a Assessments tasks like a This includes Take time because a Examples could be
word, a phrase or direct response, reading interaction and test long discussion involves speech, oral
aloud, sentence and comprehension more than one
even a sentence presentation,
dialogue completion, participant.
(greetings, small talk) storytelling, etc.
and picture cue tasks. An example could be a
and does not Non-verbal responses
Focuses on grammar, group discussion
take time except the informal
phrasal, lexical, or This type could be
monologue conversion
phonological transactional language
relationships in the or interpersonal
language used exchange
 Micro skills: Macro skills:
refer to the production of focus on fluency,
SKILLS
smaller chunks of language discourse, function, style,
such as phonemes,
OF cohesion, nonverbal
morphemes, words, SPEAKING communication, and
collocations and phrasal strategy options.
units.

 DESIGNING
ASSESSMENT TASKS:
IMITATIVE
SPEAKING
Imitative Focuses more on the pronunciation that
helps students to be more comprehensible

speaking This type is called occasional

phonologically focused repetition tasks

Repetition can be homonyms of words

, a sentence statement or question

Test example could be

PhonePass Test and computer-based test.
DESIGNING
ASSESSMENT TASKS:
INTENSIVE
SPEAKING
Intensive Students are promoted to produce short
stretches of discourse through
demonstrating
Speaking linguistic ability at specific levels.

Many tasks are cued

Part C and D of PhonePass Test fulfilled

this criteria
Examples of Intensive
Speaking Tasks:
Direct response tasks
Performance on these items is
responsive, rather than intensive

Respond with certain expected form.

Provide a little more time

Test-takers output: Form-filling, or

"oral questionnaire" (Underhill, 1987)
Read aloud test
Reading beyond the sentences could be
a paragraph

Recording the outputs makes it easy to

assess

Errors and questionable items were

noted by teachers for proper feedbacks
Students read dialogue of omitted lines by the
speaker
Sentence/dialogue
Test takers are first given one to read through
the dialogue to prepare an appropriate Completion Tasks and
response.
Oral Questionnaires
 minimal pairs

Picture-Cued
Tasks

Performance at both intensive and extensive

level
Composed of a series that tells a story or
incident
Stimulated through a more elaborate picture

comparative
Test-takers are asked to orally identify
Open-ended, not only to identify certain specific
information but also to elaborate with their own
opinion
Scoring depends on the expected performance
criteria
Must simply produce the relevant linguistic markers
Translation (of Limited
Stretches of Discourse) RESPONSIVE Question and Answer
SPEAKING

Direct approaches to creating Brief interactions with Display question;

communicative classrooms. your partner a predetermined correct
Given native language word, phrase,
Increased creativity response
or sentence and asked to translate it
Referential questions;
Advantage: control of the output
which means scoring is more easily given more opportunity to
specified produce meaningful
language in response
 Giving Instructions and Directions

To give instructions and directions you have to make sure that you know why you are asking the question.
The administrator poses the problem, and the test-taker responds.
Test-takers knowing and preparing for them in advance.
Choice of topics needs to be familiar enough.
Test-takers produce at least 5 or 6 sentences.
 Paraphrasing

Read/ hear& paraphrase limited

number of sentences
The more authentic the context is, the
more relayed the test takers are
The evidence of test takers’
conversational art in concise the input

Must clearly identify task’ objectives

 TSE
Test of Spoken English Structure
20 minute audiotape test with written, graphic, and stimuli
spoken in oral language competency for professional purposes
eg. Certifying health professional officers

Specifications
Design for eliciting oral production in various discourse than
focusing on linguistic competency in non-native speakers
Lazaraton&Wagner(1996) develop the principles to access test
takers‘ prior knowledge

Result
The validity of task between task function and test
takers’ output
 TSE
Scoring Texonomy
Scores imply to the abilities that come from effective
communication and competent performance of task

Contributing factors— Pronunciation, Grammar, and Fluency are

being decided on the overall comprehensibility

The ranges contain with task performance, function

appropriateness, and coherence

Caution
Because of the holistic of taxonomy, the institution
must provide more elaborated in scoring analysis to
encourage washback
DESIGNING ASSESSMENT
TASKS:
INTERACTIVE
SPEAKING
 Interactive Speaking Task

Interactive speaking Extensive speaking

Require a long interactive discourse Task with long duration with little interaction
Interviews, Role play, Discussion, Games Giving speech, Story telling, Explaning and translating

Differentiation is amount of interaction:

interpersonal, transactional speech event
Interview
test admin. and test takers proceed through the protocol of
questions. The data should be recorded for scoring process.

Scoring criteria can be design up to the objectives

distributing by the institutions

1. Placement interviews: Gain the quick sample from the

test takers to verify their levels in course.
2. Comprehensive interview: Predetermine the oral
production contexts
Canale’s Interview Stages

1. Warm-up : direct conversation to relax, decrease test

takers’s anxiety
2. Level check: Ask the planned question to stimulate the
responses that should come off in expected form. This is
to confirm the accuracy of the informed level of the test
takers
3. Probe: The question challenge the fullest ability of test
takers. The data will show the limitation of test taker’s
competency.
4. Wind-down: Test admin.try to help the test takers‘ minds
into normal stages, and inform the essential information
Brown’s scoring texonomy
FSI Levels
Issues on scoring system
The output from test takers likely to be susceptible or unreliable as
the pressure from the interviewing process

The ambigious distinction of score ranging such as FSI,

Brown’s

the process of 2 test takers and 1 test admin.

Advantages Disadvantages
practicality of scheduling the difficulty of equalizing the
two candidates at one outputs between candidates
time The unequal discerning of
the increase interaction comprehension and
and authenticity outputs production abilities, scoring
between candidates through the process
 Role Play
allow students to express creativity in
linguistic output within constraint of
the task
giving the prompts to test taker to
elicit certain oral abilities
scoring techniques can varies
according to testee's technique
Discussions and
Conversations informal techniques allow student to
perform genuinely and continually

such abilities will be observed;

maintenance, interrupting, clarifying,
etc.

the assessment must parallel with

task's objectives and noted testee's
details
Games
including,
1. TinkerToy - 1-2 people allowed to see the structure
and tell the others blindfolded how to recreate that
structure as accurately as possible
2. Crossword Puzzle - name the class members who
match the clues in the puzzle
3. Information Gap Grids - conduct mini-interview for
classmates to fill the boxes
4. City Maps - a city map is given to one student, who
tells their partner the directions to trace the route to
the correct final destination
 Oral Proficiency Interview
(OPI)

Originally known as Foreign Service Institute (FSI) test, is the

progressive revision from Educational Testing Service and the
American Council on Teaching Foreign Languages (ACTFL). OPI is
used across many languages around the world.

OPI is designed to assess pronunciation, fluency, grammar, sociolinguistic,

cultural knowledge, and many more. The performance will be judged by a
certified examiner on ten levels; as seen in this guidelines
ACTFL proficiency guidlines
 These guidelines offer different descriptors. Vygotskyan commented that OPI forced the
First, focus strongly on integrative abilities to test takers into a strictly controlled system with
complete the goals of the tasks rather than only the interviewer endowed. There are no
discrete skills. negotiations, practical interactions used in the
social world. The interviewees cannot predict
Second, holistic scores of ACTFL don't suit how they will deliver in real situations with a
small environments like the classroom, which native speaker.
FSI does. Bachman (1988, p. 149) also pointed out that the
validity of OPI cannot be demonstrated because
Third, the requirements of ACTFL to render the it confounds abilities with elicitation procedure
OPI is less useful in a classroom setting (Reed in the design
& Cohen, 2001)

It was emphasized that the administrators paid

and were trained to be the examiners, which
created the reliability.
Summary of ACTFL
OPI
Designing assessments: Extensive
Speaking
 Oral Presentations

The rules of assessment must include (a)

specify the criterion, (b) set appropriate
tasks, (c) elicit optimal output, and (d)
establish practical, reliable scoring
procedures.

Holistic scores obscure the variability in

across the several subcategories,
therefore is not suitable
Picture-Cued Story-Telling

providing the test takers a set of

pictures to have them tell a longer
story or description.

By providing the test material, the

examiner must be clear what skills
or abilities to assess
Retelling a Story, News Event

Test takers hear or read about news event

that needs to be retold. Which is different
from paraphrasing. The sequences and
relationships of events, expressions,
fluency, and interactions with the hearer
will be observed.
Translation (Extended Prose)

The test takers are provided with longer texts to read in

the native language and then translate into English. The
test material can be almost any kind.

Even though the translation gives the controlled set of

grammatical, vocabulary, and discourse features, the
ability of the individuals to translate them is unstable.

The score criteria should take into account the possibility of

errors unrelated to oral production ability.
Thank You for
Listening!