PRINCIPLES OF LANGUAGE ASSESSMENT
Five things
we are going to deal with this chapter are the five cardinal criteria for
“testing a test”: practicability, reliability, validity, authentic, and
washback.
PRACTICABILITY:
Taking into
account that an effective test is practical, which includes being not
expensive, staying within appropriate constraints, being easy to administer,
and having a scoring or evaluation procedure that is specific and
time-efficient. We can say that if don’t follow this suggestions seriously our
test is evidently not going to be as the name indicates, practical. I would
like to point out the fact that a test considered as such, will be really
helpful for us cause it will save time and let’s face it, we never have enough.
RELIABILITY:
For a test
to be reliable we have to consider the following possibilities:
Student-Related
Reliability, this focuses on the students’ attitude and disposition the day of
the test. Did they have a problem at home? Are they sick? Implications like
that one are relevant for the test result.
Rather
reliability: this deals with the person who scores the test and what if there
are two. It is not an easy thing to grade 40 tests when all the answers may
vary and you don’t have enough time. So, try to be as fair as you can with you
students, sometimes it is difficult because we have the bad ones that irritate
us, and the good ones that makes us feel the satisfaction of the accomplished
duty, but we have to give all a chance to show what their abilities can do.
Test
administration reliability: the environment is also an important factor at the
moment of applying a test. You have to take into account the conditions of the
classroom, the weather and noise, so your students can really take the test in
the most appropriate conditions.
Test
Reliability: Don’t forget that even the best student can be affected but simple
things as the clock ticking. So pay attention to those long exams and the
number of possible correct answers they may have. Also keep in mind that some
answers can be correct for the student depending on his own point of view.
VALIDITY
Being the
most complex criterion validity is defined as, the extent to which inferences
made from assessment results are appropriate, meaningful, and useful in terms
of the purpose of the assessment. (Gronlund, 1998, p.226) for this to be accomplished
we have to take into account these five types of evidence:
Content-
related evidence: this can be identified if you can define the achievement you
are measuring. So speaking has to be evaluated through an appropriate speaking
test, writing with writing and so on.
Criterion-
Related Evidence: it usually falls into one of two categories. One is
concurrent validity which appears when the results are supported by other
concurrent performance beyond the assessment itself. And predictive validity
which takes its importance on placement tests, admissions assessment batteries,
language aptitude tests, and so on.
Construct-Related
evidence: a construct is any theory, hypothesis, or model that attempts to
explain observed phenomena in our universe perceptions. This helps us to
identify whether the text designed was really based on the approach we wanted
it to be or not.
Consequential
Validity: this wraps up most of the things we have talked about, as considering
the accuracy in measuring intended criteria, its impact on the preparation of
test-takers, its effect on the learner and the social consequences of a test’s
interpretation and use.
Face validity:
it depends on the view of the person who takes the test and if it fulfilled all
his expectations or if on the other hand it was not what he expected and the
reason why he could not succeed.
AUTHENTICITY
We know
that a test has got authenticity if it has:
·
A
natural language usage.
·
Contextualized
items
·
Meaningful
topics
·
Thematic
organization, such as a story line or episode
·
At
least closely approximate, real world tasks.
WASHBACK
It concerns
the effects test have on instruction in terms of how students prepare for the
test and the effects on preparation for the assessment. Informal assessment is
more related to interactive feedback. Remember
that if you just give a number to the assessment or the test you are doing it
wrong!
As we can
clearly see, to prepare a good test we have to take into account many things
that may sound a bit out of range but I am pretty sure that after having read
all this factors, at least, we will think it twice before we apply a test to
our students and of course at the moment of giving it a grade or the feedback
the student needs.