2019 Fiscal Year Research-status Report
Developing a diagnostic language assessment of Japanese EFL learners' English language email writing
| Project/Area Number |
19K00826
|
|---|
| Research Institution | The University of Aizu |
|---|
Principal Investigator |
Nicholas Allan 会津大学, コンピュータ理工学部, 准教授 (50811048)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
BLAKE John 会津大学, コンピュータ理工学部, 准教授 (80635954)
Mozgovoy Maxim 会津大学, コンピュータ理工学部, 准教授 (60571776)
|
|---|
| Project Period (FY) |
2019-04-01 – 2022-03-31
|
| Keywords | pragmatics / assessment / corpus |
|---|
| Outline of Annual Research Achievements |
Tasks for eliciting email text data were created, based on questionnaire data from a sample of the student population. Data were ranked in terms of frequency, with the most frequent serving as templates for tasks. Draft tasks were validated by expert English users for, and then administered online to 426 student participants. From the participants, approximately 1,500 texts were elicited and collated into a corpus of learner emails. A sample of the corpus was subject to preliminary analysis, identifying pragmatic failure relating to politeness, formality and email structure.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
The project is progressing according to the timetable set out in the application. Tasks have been created, data elicited and collated into a corpus, and a preliminary analysis has been conducted.
|
| Strategy for Future Research Activity |
In the next phase of the project, the full corpus dataset will be manually annotated by expert English users for specific instances of pragmatic failure. Inter-rater reliability will be checked to ensure a reasonable level of agreement between annotators. A prototype of the diagnostic learning assessment online tool will be developed, using the corpus to help automatically identify failure. This prototype will be piloted on a small group of participants to check effectiveness.
|
| Causes of Carryover |
Our original intention was for expert English users to be paid to annotate the full data set last year, but this work will be done this year instead.
|
