Professional Documents
Culture Documents
Deat Lecture 02 - Dr. Lounaouci 2023-2024
Deat Lecture 02 - Dr. Lounaouci 2023-2024
Deat Lecture 02 - Dr. Lounaouci 2023-2024
In order for students’ different levels of cognitive (i.e., thinking) skills to be measured,
teachers are required to espouse a multiplicity of assessment methods and techniques when it comes
to designing test questions and items. According to Brown and Hudson (1998), such tasks, whose
evaluation is carried according to the five principles of assessment delineated in lecture 01, include
selected response assessment, constructed-response assessment, and personal response assessment.
However, in this lecture, we are only going to focus on the first two methods, namely selected and
constructed response assessments. Besides, having mentioned the phrase ‘different levels of
cognitive skills’ in relation to the two assessment methods, this lecture also aims at correlating
learning objectives and test items to Bloom’s Taxonomy since different assessment methods and
test items demand distinct levels of thinking skills (i.e., lower-order vs. higher-order thinking
skills), an issue that will be addressed as part of both selected and constructed-response assessment
methods.
1. Selected-response assessment
In selected-response assessment, students are not required to produce any language; they only
need to ‘choose’ or sometimes ‘guess’ the right response from among a set of ‘limited’ options.
This, therefore, means that these kinds of assessments are most appropriate for measuring receptive
skills like listening and reading.
Teachers should be able to construct quizzes that target higher-level thinking skills (consistent
with the application, analysis, evaluation, and creation/creativity levels of Bloom’s taxonomy), and
they should evaluate their instruments by conducting item analyses.
Nilson (2016, p.291) notes that these tests, which are easily ‘designable’, ‘administrable’
‘scorable’, are good for assessing students’ ability to remember and understand (and probably
apply) course concepts and materials, but cannot “measure students’ abilities to create, organise,
communicate, define problems, or conduct research”, hence lower-order thinking skills.
1
Selected response tasks are known to be highly practical as they do not demand much time
and energy to design, administer, and correct. One must, however, point out that multiple choice
questions present an exception in that they are difficult to design. Finding the right distractors that
are neither too easy nor too difficult is challenging for teachers.
Selected response tasks are also highly reliable since only one answer is possible and
answers are either correct or incorrect. Raters score these types of tasks objectively without the
interference of any judgment.
Validity can only be achieved if the nature of the task matches the objectives of the course
and the nature of the skills assessed.
Selected response tasks lack authenticity as the language used is unnatural and
decontextualised, and the items are isolated. Similarly, the tasks themselves lack authenticity
because they do not mirror real-world tasks performed by native speakers.
Selected response tasks do not offer much washback. Language-wise, learners are not
required to perform, create, or produce anything. Drawing upon Bloom’s Taxonomy, the selected-
tasks also do not require higher-order thinking or problem solving skills. This being said, some
positive washback can be achieved if and only if the tasks are relevant to what has been taught
(validity) and if after returning tests, teachers provide feedback and comments that help learners
identify their mistakes and correct them.
2. Constructed-response assessment
These activities lack practicality as they are time and energy consuming in their design,
administration, and scoring.
They also lack reliability as the answers are open-ended. Unlike selected response
assessments that present only one possible right answer, their constructed-response counterparts
2
offer a range of possibilities leaving place for debate about answers correctness. This, of course,
leads to intra- rater reliability and inter- rater reliability issues.
Validity in this type of assessment depends on the objectives of the course and the nature of
the skills evaluated. A test which aims at evaluating the reading comprehension skill should make
use of information transfer activities to achieve high level validity. If reading comprehension
questions that require written production of language are used instead, validity is affected. It will
not be clear in this situation if the scores obtained reflect students’ reading ability or writing ability.
Authenticity is better achieved with constructed response assessments than with selected-
response assessments. The language is more natural and is presented around a given theme or topic.
The tasks are also more authentic as native speakers are involved in producing pieces of writing and
taking part in conversations and interviews.
Washback is also more important since this type of assessment promotes learning and
contributes to the development of learners’ language. Learners receive qualitative feedback and
constructive criticism that can help them improve their performance. When assessment is relevant
to the content of the course, students are urged to prepare for the test, and review their lessons
individually or in groups. This increases their opportunities of learning.
1. What is the link between teachers’ perceptions and practices of language assessment?
2. How do you (Master II students) perceive language assessments and what factors affect such
perceptions?
******************************************************************************
References
Brown, J. D., & Hudson, T. (1998). The alternatives in language assessment. TESOL Quarterly, 32
(4), 653-675.
Nilson, L. B. (2016). Teaching at its Best: A Research-Based Resource for College Instructors (4th
ed.). San Francisco, CA: Jossey-Bass.