Upload
ayu-delisa-putri
View
217
Download
0
Embed Size (px)
Citation preview
Validity
By the far most complex criterion of an effective test-and arguably the most importand
principle is validIty,” the extend towich imperence made from essesment results are
appropiate,meaningfull,and usefull in terms of porpuse of the assesment. A valid test of reading
ability actually measures reading ability not 2!2 vision,nor previos "nowladge in a sub#ect,
nor some other variable of $uestion rrelevance. %o measure writing ability, one ma"e as" the
students to write as many words as they can in &' minutes, then simply count the word for the
final score. (uch a test would be easy to administer)practical*and the scoring $uite depandable
)realible*. But it would not constitute a valid test of writing ability without some considaration of
ideas,among other factors.
+ow is the validaty of a test established %here is no final, absolute measure of validity,
but several different "ind of evidence may be invo"ed in support. In some cases, it may be
appropiate to examine the extence to wich a test calls for ferformance that mathes that of the
course or unit of the study being tested. In other cases we may be concerned with how well a test
determines wheater or not students have reached an established set of goals or level of
competence. (tatistical correation with other releted but independent measure is onother widely
accepted from of evidence. ther concers about a test/s validity may focus on the conse$uences
0beyond measuring the criteria themselves-of a test or even the test-ta"er/s perception of
validity. 1e will loo" at the these five types of evidence bellow
&. content related-evidenceIf a test actually samples the sub#ect matter about wich conclusions are to be
drawn, and if it re$uires the test-ta"er/s to nperform the behavior that ios being measured,
it can be claim content -related evidence of validaty. ften popularly referred to as
content validaty. 3ou can usually identify content-related evidance observationally if you
can clearly define the achievment that you are measuring.
nother way of understanding content validaty is to consider the difference
between direct and indirect testing. 4irect testing involves the test-ta"er/s in actually
performing the target tas". In and indirect test, learners are not performing the tas" itself
but rather a tas" that is related in some way. 5or example if you intent to test the
learner/s oral production of syllable stress and your tas" test is to have learners
mar"s)with writtent accent mar"s* stressed syllable in a list of written word, you
could,with a stretch of logic, argue that you are indirectly testing their oral production. A
direct test of syllable production would have ton re$uire that students actually produce
target words orally.%he most feasible rule of thumb for echieving content validaty in classroom assesment is
to test performance directly. 6onsider, for example a listening7spea"ing class that is doing
a unit on greating and exchanges that includes discourse for as"ing for personal
information) name, address,hobbies, etc* with some form-focus on the verb to be,
personal pronouns and $uestion formation. %he test on the unit should include all of the
actual prformance of listening and spea"ing.
2. 6riterion-8elated 9vidance
A second form of evidance of the validaty a test may be found in what is called criterion-
related evidance, also referred ton as criterion-related validaty,or the extent to which the
criterion of the test actually been reached.
6riterion related evidance usually falls into one of two categories concurent and
predictive validaty. A test has concurrent validaty if its results are supported by other
concurrent performance beyond assessment itself.forexample the validaty of a high score
on the exam of a foreign language courses will be substantiated by actual profiency in the
language. %he predictive validaty of an assessment become importand
In the case of placement tests,admission assessment batteries, language aptitude
test, and the li"e.the assessment criterion in such cases is not to measure concurrent
ability but to asses )and predict* a test-ts"er/d li"elihood of future succes.:. 6onstruct-8elated 9vidance
A third "ind of evidance that can support validaty, but that does not play as large a role
for classroom teachers, is construct-related validaty, commonly reffered to as construct
validity. A construct ia any theory, hyphotesis or model that attempts to explain observed
phenomenon in our universe of perception. 6onstructs may or vmay not be directly or
empirically measured 0their verification often re$uaire inferential data. ;rofiency and
communicative competence are linguistic constructs , self-esteem and motivation are
psyhologycal constructs.
<. 6onse$uncetial =alidity
As wall as the above three widely accepted forms of evidance that may be
introduceed to support validay of an assessment, two other categories may be of some
interest and untility in your own $uest for validating classroom tests., among others,
underscore the potential importance of the conse$uence of using an assesment.
6onse$uance validity uncompassed all the conse$uence of a test, including such
considarations as its accuracy in measuring intended criteria , its impact on the
preparation of test-ta"er/s, its effect on the "learners, and the intended and unintended*
sicial se$uences of a test/s interpretation and use. In other word conse$untial validaty is
+ow well use of assessment results accomplishes intended purposes and avoids
unintended effect.
'. 5ace =alidaty
An importand facet of cconse$uential validaty is the extent to wich students view the
assessment as fair,relevant, and usefull for improving learning. or what is popularly
"nown as face validity . face validity referst to the degre to wich a test loo"s right. And
appears to measure the "nowladge or abilities it claims to measure, based on the
sub#ective #udgment of the examines who ta"e it, the administrative personnel who
decide in its use, and others psychometrically unsophisticated observers.
(ome time student don/t what is beingn tested when they ta"le a test. %hey may fell,
for a variaty of reason, that a test is not testing what it is supposed to test. 5ace validaty
means that the student perceive the test to be valid. 5ace validity It can be empirically
tested by a teacher or even by a testing expert because it is based on the sub#ective
#udgment of the examinees who ta"e it.
Authencity
A fourth ma#or principle of language testing is authenticity, a concept that is a litle
slippery to define, especially whitin yhe artr and science of evaluating and designing tests.
Bachman and ;almer Aunthenticitydefine authenticity as a degree of correspondence of the
chracteristics of a given language test tas" to the futures of a target language tas" and then
suggest an agenda for identifiying those target language tas"s and for tranforming them in to
valid test items.
In a test authenticity may be present in the following ways
• the language in the test is as natural as possible.
• Items are contextuali>ed rather than isolation.
• %opict are meaningfull)relevant?, interesting* for the learner.
• (ome thematic organi>ation to items isprovided, such as through a story line or episode.
• %as" represent, or closely approximatye, real-world tas"s.