language. Since this PhD dissertation was written in English, the researcher
translated the original recording directly into English. Again, in order to gain a
strong“sense of event”, the researcher listened to the recording from time to
time, especially in cases where the participant’s report was vague in meaning or
where the participant’s intention was not directly expressed. The only criterion
is that the translated transcript could be as faithful to the original as possible.
b. punctuation. Transcripts were punctuated as sentence units and a new sentence
will be shifted to a new line. A unit for analysis can be a sentence, a clause or
even a phrase (Green 1998: 73), because it varies depending upon the coding
scheme. For the current study, a unit for analysis is usually a sentence or a
clause with verbal protocols containing intact meaningful cognitive operations.
c. quotation. The words quoted from the outline of the mini-lecture were quoted
with quotation marks in order to differentiate from the verbal protocols. For
example,“think before write”as a mentioned piece of information from the
given prompts in the gap-filling task is quoted because participants would use
words such as“think”or“write”very often in their verbal protocols.
d. additional information. Additional information, either a longer pause in the
verbal protocol or paralinguistic features were indicated in the parentheses right
after the sentence; for example, (pause 5 s) means the participant paused for 5 s
or (rising tone) means the participant was using a rising tone in the recording,
etc.
e. a proportion of transcripts were checked against the original recording by a
second coder so that the reliability of the transcription is guaranteed.
The transcripts for three phases listed above across all the 16 participants amount
to 36301 English words.
6.5.1 Segmentation....................................
The transcription of the audio data isfirst divided into three source folders, i.e., the
folder of Phase 1 task-tackling, the folder of Phase 2 retelling and the folder of
Phase 3 interview.
All the transcription was segmented into sentences or clauses that accommodate
intact propositional units. In order to make sure that all the meaningful sentences or
clauses are coded into their corresponding exact nodes (cognitive
operations/processes), sentences or clauses containing more than one cognitive
operation would not be combined in one paragraph and a new line was used for a
new coded cognitive operation. If a paragraph with several coded cognitive oper-
ations was not divided into different lines, it would make the intra-coder and
inter-coder reliability check more inconvenient.
80 6 Employing the Think-Alound Method...