kind of interaction? Traditionally, selected response format such as MCQ is pre-
ferred in high-stakes tests with a prominent advantage of convenient scoring and it
is an indirect test measure, though. But due to deficiencies of selected response
format, more constructed formats have been used in the current academic listening
tests, such as note-completion, summarization, etc. When constructed format as a
summary cloze is handled, the task targets, or gap types concerning which part of
the content to be tested should be calibrated toward certain cognitive processes.
Task setting includes weighting, e.g., weighting of different tasks and the
weighting of listening subtest in a comprehensive EAP test, time constraints and
knowledge of criteria. Knowledge of criteria refers to marking criteria. Especially
for constructed response format, how marking criteria are implemented need to be
made transparent to the test-takers.
Finally,task demandsencompass aspects of knowledge resources and skills
needed for lecture comprehension. The aspects of knowledge resources include
linguistic knowledge, pragmatic knowledge, cultural knowledge, content knowl-
edge, etc. Skills required for lecture comprehension include identifying key ideas
and topical areas, grasping discourse structures, taking notes, etc. The reason why
skills are put in the task domain is thatfirst, the task domain departs from the
test-designer’s or teacher’s point of view and conventionally, skills needed for a
specific test are mandated by the test-developers. In order to match skills tested in a
lecture comprehension task and skills really needed in the target-language domain,
we definitely call for more needs analysis research to address this issue. Second,
based on Bachman’s (1996: 75–76) argument about language skills, skills can be
seen as a combination of“language ability and task characteristics”. He didn’t
consider language skills to be part of language ability, but to be the“contextualized
realization of the ability”, to use language in specific language use tasks (Bachman
1996: 75–76). The essence of Bachman’s argument is that skills should be con-
sidered as specific activities or tasks in which langue is used with purpose. Thus,
academic listening can be defined as an academic language use task that involves
the activity of listening and be described in terms of task characteristics combined
with components of language abilities. As a working definition of skills here in the
framework, they are not as broad as conventional language skills as listening,
speaking, etc. but relevant components of language ability needed for an academic
language use task, more specifically, for academic lecture comprehension. Thus, if
we still assign skills needed for academic listening to our framework, it should
belong to the domain of task characteristics, namely, the task domain.
Competence domainis the research focus here, with the bold font to show it is
the most salient part of the current construct framework, represented by test-takers’
cognitive processes. If test-designers depart from the task domain in the process of
a test development, their target should be the test-takers’real cognitive behavior.
Bachman and Palmer (2010: 57) also integrated cognitive processes or strategies in
language ability:
40 4 Task Characteristics of TEM 8 Mini-Lecture...