task demands, but to what extent the assumed task demands can be actualized in the
real test-taking cognitive process must be supported by qualitative data. The current
TAP data prove that gaps targeting details successfully elicit test-takers’lower-level
processes such as decoding, lexical search, parsing, and furthermore, there are cases
providing signs of meaning construction process. For example, participants tried to
interpret meaning of concepts in their own way and build connections between
propositions in order to attain semantic coherence, etc.
8.3.3 Inference Gaps
This type of gaps aim to elicit test-takers’inference of the given information.
Usually, the word/words to befilled in the gap does/do not directly appear in the
mini-lecture. Obviously, this type of gaps could be highly cognitively demanding.
Without understanding the relevant details of the mini-lecture, the test-takers might
feel at a loss while tackling these gaps.
Phase 1 protocols have proved that participants reached the level of meaning
construction with evidence of a full range of cognitive operations employed to help
build meaning in the context, including contextualizing notes, inferring meaning
from notes or background information, interpreting concepts in one’s own way, etc.
The protocols also foreground the importance of identifying and selecting key
words and writing them down in notes for later retrieval and the importance of
building links, including links between propositional units and links between input
and schema. However, thesefindings cannot be triangulated with any extracted
main component from the questionnaire data. When we trace back to the ques-
tionnaire data, we mayfind that items with reference to word meaning inference, or
connotation of relevant lecture content are not closely related to the main compo-
nents, and they were accidentally grouped into either the component of decoding or
of micro discourse construction.
8.3.4 Summary Gaps
This type of gap requires test-takers to summarize a relevant block of information.
The target of this gap type is to attain the test-takers’discourse construction pro-
cess. Based on the Phase 1 protocol analysis, this type of gaps is very difficult to
test-takers and also poses high cognitive demand. The difficulty mainly lies in two
aspects: 1. how to select a complete relevant sequence of propositions for summary;
- how to generate the given sequence of propositions or reconstruct the sequence
of propositions in one’s own way. So only those participants who can successfully
solve these two problems might work out the correct answer. In Hansen’s study
(1994) focused on the correlation between question type and students’performance
on authentic lecture comprehension, segments of actual university lectures were
134 8 Linking Task Demands, Cognitive Processes...