2018 Fiscal Year Research-status Report
Employing corpus-based Word Sketch collocation summaries as an L2 writing reference tool: The impact on lexical accuracy and diversity of Japanese EFL compositions
| Project/Area Number |
17K02927
|
|---|
| Research Institution | Kobe University |
|---|
Principal Investigator |
QUINN Cynthia 神戸大学, 国際文化学研究科, 特任准教授 (00368474)
|
|---|
| Project Period (FY) |
2017-04-01 – 2021-03-31
|
| Keywords | L2 writing / Corpus referencing / Lexical accuracy / Word Sketch |
|---|
| Outline of Annual Research Achievements |
In H30 the aims of the study were streamlined in order to (1) narrow the research focus to lexical accuracy only rather than investigating both accuracy and diversity; and (2) to collect data on the learners' decision-making processes as they employ Word Sketch on the corpus. To accomplish this, a second round of data was collected, which improved upon the pilot study in H29. Specifically, both the experimental task and process were revised to be consistent with this narrowed research aim. In addition, participant discussions were audio recorded, which are a new data source from last year's pilot and will allow for better triangulation of the data overall. Finally, the researcher gave presentations at three international conference last year.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
In H30, another round of data was collected from 28 participants. Compared to the pilot study conducted in H29, two major improvements were made this year: (1) the writing task was streamlined to focus on lexical accuracy with students correcting an of average 15 teacher-designated lexical errors with Word Sketch; (2) participants worked in pairs and discussed their corrections aloud so that their decision-making process for each correction could be verbalized and recorded during the task. As with the pilot study in H29, the participants' error correction processes and Word Sketch referencing attempts were computer-screen recorded. Aside from data collection, transcription of the participants' discussion data has been underway since the latter part of H30.
|
| Strategy for Future Research Activity |
Data collection is now complete, so from this point forward efforts will center around data analysis. This year's work will involve transcribing the audio data of the participants' discussions, which includes 14 ninety-minute audio files that represent data from 14 pairs or 28 participants. Research assistants have been hired to transcribe these files (in Japanese), and then the researcher and one other research assistant will work together to translate key parts of the transcriptions into English. Besides the audio data, the participants' essays and video data must be analyzed, which will be done this year by the researcher and one other research assistant.
|
| Causes of Carryover |
For data analysis this year, data management software and data storage devices will be purchased, since the audio and video files amount to a large amount of data. One continued expense this year will be the research assistants' fees, as transcribing the data is extremely time-consuming. Otherwise, travel expenses will be incurred for conference attendance and possibly research trips in order to collaborate with other corpus linguists and software developers who can assist with analyzing the large amount of data generated from this study.
|
Research Products
(3 results)
