Abstract Transcription is an example of a task that requires an intensive interaction with a sound recording. This paper describes a controlled experiment conducted to determine the differences between native and non‐native speakers strategies for transcribing dialogues, from the simple to the complex. The results of the analysis confirmed there were significant differences amongst three categories of the users’ interactions with a computerized transcription system: 1) the number of times a recording segment was replayed, 2) the number of text segments used in transcribing the dialogue, and 3) the number of times a portion of the text was deleted. Moreover, the results suggest that the difficulty of the task in combination with the user interactions, such as the number of replays and the number of text deletions, is a predictor of the language proficiency. These findings can be used to develop adaptive user interfaces, to determine an optimal sound quality, as well as adjusting learning and transcription ...
[1]
Mari Noda,et al.
Japanese, the Spoken Language
,
1987
.
[2]
Martha E. Crosby,et al.
Individual Differences in the Use of Hyper/Multimedia by Undergraduate Students at the University of Hawaii at Manoa
,
1993,
HCI.
[3]
J. Hulstijn.
When Do Foreign‐Language Readers Look Up the Meaning of Unfamiliar Words? The Influence of Task and Learner Variables
,
1993
.
[4]
E. Barrett,et al.
Textual intervention, collaboration, and the online environment
,
1989
.
[5]
Martha E. Crosby,et al.
HYPERMEDIA AS A FACILITATOR FOR RETENTION A CASE STUDY USING KANJI CITY
,
1994
.
[6]
B. Curtis,et al.
Measurement and experimentation in software engineering
,
1980,
Proceedings of the IEEE.