Unit 7 & Unit 8

Scoring ferformance tests & interluctor training and

behaviour

By :
Defri Sajo
Adam Rahman Arif
 Scoring ferformance tests
Assessing writing and speaking samples has
always been considered important for two main
reasons. Firstly, the view has often been
expressed that the best way of predicting writing
or speaking performance in ‘the real world’ is to
getstudents to write or speak in a test. Secondly,
it has been assumed that only by including
writing and speaking in a test will the test have
the required washback on the classroom
 introduction
The first problem that we encounter is defining
‘the real world’, and trying to recreate that in
the test. Then comes the problem of how to
score the limited sample that we are able to
collect in the test, so that the score carries
meaning for the target domain and is relevant
to any instructional context that is preparing
learners to become competent communicators
within that domain.
 ESL Witing : a different context
ESL writers often acquire different components of
written control at different rates. Every instructor
of second language writers has encountered those
students who have fluency without accuracy and
those with accuracy but little fluency. We also
sometimes see writers who have mastered a wide
vocabulary but markedly less syntactic control;
or syntactic control not matchedby rhetorical
control; and so on.
even more important, the chances of significant
improvement in writing are
greater for ESL writers than for most L1 writers.
By the time they reach college most first
language writers have achieved most of their
writing skills.
 Advantages of multiple trait instruments
 Salience.
By ‘salience’ I mean that the writing qualities evaluated, and the kinds of writing
samples collected are those that have been found appropriate in the context
where the assessment takes place .
 Reality and community
One advantage of multiple trait scoring is that, once the initial process of data
collection, prompt and rating guide development and validation is complete, and
readers have been trained, readers’ judgments are linked to something
which is at once external and internal.
 Increased information.
The chief advantage of multiple trait scoring combined with
profile reporting, and particularly important when working with ESL writers, is
that the judgments made by assessment readers can be translated into
information which can be shared with the writers, their academic advisors and
other concerned parties.
 If a writer’s overall performance puts her into the category
of those who will receive
special courses or other special services, by looking inside
the information provided by
the multiple trait instrument, that is, by looking at the score
profile, the writer, the
class teacher, and the program administrator can make good
decisions about which
course offering or other kind of service would most help
this individual writer make
progress.
 Interluctor training and behaviour
We now know that it is important to train the people
who are going to be involved in test administration.
At its most basic this involves the logistics of simply
getting things done, such as providing the right-size
chairs or making sure that the space for testing is
properly organized (see Fulcher, 2003a: 152–153),
through to ensuring
that raters or interlocutors in a speaking test employ
the behaviours that are needed
to elicit a ratable speech sample.
 Methodology of the discouse analysis
Given the focus of the assessment on candidates’ communicative
effectiveness, the perspective taken within the analysis is also
on communication, on the interplay
between interviewer and candidate. As an analysis of interviewer
moves alone would reveal little about their impact on candidate
performance, the analysis consists not
simply of a count of prespecified interviewer behaviours, but of a
sequential analysis of the talk as a whole. This will allow us to
ascertain not only the ways in which interviewer behaviour
differs but also how these differences affect the quality of the
candidate’s talk and construct different pictures of her
proficiency.
 conclusion
It appears from this study that standard approaches to the training
and re-accreditation of examiners may be less than satisfactory.
Interviewer training has generally tended to be somewhat
overlooked in relation to rater training, with interviewer
behaviour rarely being scrutinized once initial training is
completed. With the emphasis in tests of second language
proficiency being increasingly on relativelyunstructured
naturalistic interactions, however thorough the initial training,
it is incumbent on test administrators to ensure that
interviewers’ styles do not over timebecome so diverse that
they present different levels of challenge.
 Thanks
Quotes today:
“Stay positife, attitude is everything.”

@ketikanGabut