Traditional vs.
Alternative Assessment
Traditional Assessment Alternative Assessment
One-shot standardized exams continuous long-term assessment
Timed m/c format untimed free-response format
Decontextualized test items contextualized communicative tasks
Scores suffice for feedback individual feedback and washback
Norm referenced criterion referenced
Focus on right answers open-ended creative answers
Summative formative
Product-oriented process-oriented
Non interactive interactive
Extrinsic motivation intrinsic motivation
Purposes of lg tests
Diagnosis, accountability & Feedback
To pinpoint strengths and weaknesses in the learned abilities of individuals.
Such tests provide information to:
students
teachers
administrators to make learning process more effective.
They can be theory or syllabus-based.
Accountability: the extent to which we have effectively and efficiently
discharged responsibility. Without it students might have passed several
conversational courses, but still unable to put the knowledge into use.
Feedback: information provided to teachers, students, and other interested
persons about the results of an educational program.
Modifications in the program can be made based on the feedback.
Screening & Selection (entrance, readiness)
Who should be allowed to participate in a given program when there are more
applicants than the space available. In other words, who can benefit from the
program more, to attain mastery of lg.
Readiness test: to determine whether or not students are ready for instruction.
Examples: scores of previous term, academic aptitude
Placement
Tests are used to identify a particular performance level of the learner
and place him at an appropriate level of instruction. It is a sort of
screening as well.
Such tests normally measure students’ lg abilities.
The content of such tests depend on:
a theory of lg proficiency (if there are large number of students with
various backgrounds)
learning objectives of the syllabus (if they are clearly defined)
Placement test can be regarded as a broad-band diagnostic test,
distinguishing weak from strong students to provide them with activities
at appropriate level.
Achievement tests
They are directly related to classroom lessons, units, or even total
curriculum. They are limited to particular material addressed in a
curriculum within a particular time frame.
Primary purpose: to determine whether course objectives have been
met and appropriate skills are acquired at the end of instruction.
They are often summative, but can also be formative providing
washback about the quality of the learner’s performance during the
course.
Specifications of achievement tests depend
on:
1. the objectives of the course
2. the importance of each objective
3. the tasks used in the course
4. practicality issues: time frame of the test
Proficiency tests
To test global competence in a lg.
They are always summative & norm-referenced. They
provide results in the form of single scores.
Their interpretation depends on norms and percentile ranks.
So they do not provide diagnostic feedback.
lg aptitude tests
They predict a person’s success prior to exposure to
the L2.
It measures general ability to learn a foreign
language and its ultimate success.
Vocab tests are effective aptitude tests, perhaps
because they correlate with intelligence tests.
Direct vs. Indirect Tests
Some tests like ratings of lg use in real and uncontrived
communication situations are testing lg performance
directly.
Some tests like multiple-choice recognition tap lg
performance indirectly.
Therefore, most of the tests lie somewhere on a continuum
from natural-situational to unnatural-contrived. E.g.,
interview is more direct than cloze. contextualized vocab is
more natural than synonym-matching test.
CRT (edumetric, common metric),
The test must match the teaching objectives. Scores are reported and
interpreted based on a specific content domain or criterion level of
performance.
A criterion or cut-off score is set in advance (usually 80 to 90% of the
total possible score).
Those who do not meet the criterion should repeat the course. Students
are not compared. Their mastery level is important.
Advantages:
to clarify objectives: to see whether they have been met.
useful for small/unique groups for whom norms are not available.
since content is predetermined, test anxiety is reduced.
Disadvantages
objectives are often too limited
students cannot compare themselves to a criterion
(norm)
techniques of estimating their reliability and
validity have just been developed, so not clear
whether they are V & R or not.
NRT (psychometric) must have been previously administered to a large
sample of people from the target population (1000 or more). Acceptable
standards are then determined. Normative interpretations of test results
are made.
If NRTs are perfectly designed the scores will be distributed normally
on the curve.
Failure of large number of students results in the revision of test not
revision of program or dismissal of the teacher.
Advantages
comparison can be made with the performance of large population of
students.
since R & V can be estimated there is more confidence with such tests.
since standards are determined based on comparison with a large group
they are fairer and less arbitrary than that of CRT.
more profound comparative information is provided than pass/fail
information in CRT.
Disadvantages
they are only valid with the population on which they have been normed.
Norms change with time like the characteristics of population.
results cannot be matched with instructional objectives, because they
are not developed based on a particular program.
debilitating test anxiety is high.
since average score is important there is insensitivity to fluctuation in
individual.
How to classify various tests?
