2006 Fiscal Year Final Research Report Summary
Fundamental Study on Second Language Writing in Japanese : What Is Good Writing?
| Project/Area Number |
16520321
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Japanese language education
|
|---|
| Research Institution | The University of Electro-Communications |
|---|
Principal Investigator |
TANAKA Mari Center for International Programs and Exchange, Associate Professor, 国際交流推進センター, 助教授 (20217079)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
NAGASAKA Akemi Keisen University, Faculty of Humanities, Professor, 人文学部, 教授 (50207990)
|
|---|
| Project Period (FY) |
2004 – 2006
|
| Keywords | Japanese Second Language Education / Writing / Second Language Writing / Scoring rubric / Multiple-trait scoring / Prompt / Rater / Scorer / Raters' / Scorers' Meeting |
|---|
| Research Abstract |
This research aims to investigate the elements that make good writing in Japanese and develop a new scoring rubric for academic writing in L2 Japanese. To date a standardized scoring rubric in Japanese language education does not exist. We believe that it is meaningful to develop a set of "Assessment Guidelines" based on writing research in English language education.
First, we conducted a review of writing evaluation studies and decided to use multiple-trait scoring for our new rating scale. Multiple-trait scoring meets the requirements of this study because there is no weighting attached to the traits, which can be selected and combined to suit the purpose. To make learners realize the importance of traits such as "purpose of writing" and awareness of the target "reader," we developed a scoring rubric which contains five traits, "Aim/Content," "Organization/Cohesion," "Reader," "Language use A : Accuracy," and "Language use B : Appropriateness," with 7 levels from 0 to 6. As for prompts, to make the evaluation of the two traits "Purpose" and "Reader" clearer, we added instructions and commentary on the process of evaluation.
A scoring scale should be user-friendly and also have high reliability and validity. We asked learners to do self-assessment using our scale and held a writing evaluation workshop for 20 Japanese language teachers after which 8 teachers were asked to score two sets of 26 samples of argumentative and expository essays. The results showed that there was a high coefficient alpha of 0.85 to 0.93 for all traits. The correlation between the prompts was high for the trait "Language use," which suggests that differences between prompts do not influence the evaluation of linguistic ability. The evaluators also pointed out at feedback sessions that the writer's language ability may have an effect on "Aim/Content" and that a scorer's background may influence his/her assessment.
|
