participants mentioned in their verbal protocols. It substantially demonstrates stu-
dents rely heavily on their notes in tackling the gap-filling task. Actually, since a
mini-lecture lasts 7 to 8 min with around 900 words, it already exceeds the normal
capacity of working memory. Students must refer to their notes to complete the
task. Students also reported in their TAPs that the words they had written down in
notes were more reliable tofill in the blanks than inferences emerging from their
own mind.
We must note that decoding and selection process occupies more than half of the
test-takers’reported cognitive operations and test-takers employ sub-processes such
as reading notes, memorizing words (14 sources, 87 references), reading the task
(13 sources, 59 references) and parsing the task (13 sources, 35 references) quite
unanimously. Self-evidently, decoding is thefirst process in listening comprehen-
sion. However, since the test-takers were completing the task after the mini-lecture,
the decoding and selection attention process here does not refer to test-takers’
online decoding of the mini-lecture, but rather the selective decoding of the
information given in the gap-filling task, notes they have written and the retained
information units retrievable in their mind. So, these three sources of information
largely shape their decoding and selective attention process. Furthermore, selective
attention here has two layers of meaning. First, the notes written down and infor-
mation units retained in the working memory are already the selected information
from the original lecture; second, while tackling the task, test-takers need to select
key relevant words from the “selected information”, i.e., they need to select
information for the second time.
8.5.2.2 Meaning and Discourse Construction
Typical meaning-building cognitive operations such asinterpreting,inferringwere
most frequently reported, which indicates high interactiveness between the task
targets and cognitive processes. The meaning and discourse construction process is
not as systematic as the decoding and selective attention process. Participants resort
to their schema to make sense of the new information and the task as well.
Therefore, the cognitive operations participants employed vary with individuals.
The unanimous use of interpreting (13 sources, 34 references) is very representative
of academic listening. Students often reported their interpretation of a new concept.
For example, a participant reported in TAP that“I think the word powerful is in
contrast to less common expression”in order to interpret the meaning of word
“powerful”. As for making inference, participants not only inferred from their own
notes, but also actively interacted with the task itself, trying to infer information
from the given information of the task. For example, a participant reported:“I guess
I might write down‘common expressions’; this blank should be in contrast to‘less
common expression’”. So the given information in the outline becomes a semantic
hint for the test-takers. Summarizing is not as frequently used by test-takers as
interpreting and inferring. Summary gaps are hard to design. And sometimes, the
summary words are already given in the lecture and hence the targeted cognitive
8.5 Localized Lecture Comprehension... 139