| Project/Area Number |
11680280
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
教科教育
|
|---|
| Research Institution | Seijo University |
|---|
Principal Investigator |
KUBOTA Mikio SEIJO University, LITETATURE, ASSOCITATE PROFESSOR, 文芸学部, 助教授 (60259182)
|
|---|
| Project Period (FY) |
1999 – 2000
|
| Project Status |
Completed (Fiscal Year 2001)
|
| Budget Amount *help |
¥2,200,000 (Direct Cost: ¥2,200,000)
Fiscal Year 2000: ¥600,000 (Direct Cost: ¥600,000)
Fiscal Year 1999: ¥1,600,000 (Direct Cost: ¥1,600,000)
|
|---|
| Keywords | language information / item respove heavy / insmental effect / comuuterized adaptive test / grammaticality judgement test / item bank / restructuring / dative verh / 個人適応型コンピュータ・テスト / 個人能力推定値 / 項目難易度パラメータ |
|---|
| Research Abstract |
The 1999 study investigated(1) what type of instruction may be most effective in enabling EFL learners (re) formulate their grammar on spontaneous production tests and(2) whether statistical differences of the results may be observed between the classical test theory(CTT) and the item response theory (IRT). 60 Japanese EFL university students were asked to describe pictures spontaneously in either oral or written modes. The four classes were assigned to four different experimental groups(Groups A, B, C, and D) according to the type of treatment. Results showed that(1)Group C proved to be more effective for one month than the other treatments. (2) When the learners received positive input, the provision of explicit metalinguistic information was more effective than that of repetition. (3) When positive input was given with repetition, explicit metalinguistic in formation was not indispensable to learners' (re)structuring. (4) All four experimental groups gained measurable instructional
… More
benefit. The learners in Group D had more opportunities and time for monitoring their performance in writing than in speaking. (5) The test modality (oral vs written test)did not cause any statistical difference. (6)No statistically significant difference was discovered between CTT and IRT. One can use IRT models practically on an equal footing with CTT in classroom settings.
The 2000 study investigated whether a computerized adaptive test(CAT) would successfully apply to practical language tests in EFL environments, and examined the similarities differences in person ability estimation between CAT and a pencil-and-paper test(PPT). Japanese university first-year students participated in the study. The results showed CAT estimated similar person ability parameters as PPT, that the number of items given to the subjects became smaller in CAT than PPT, and that CAT reliably measured learners' English proficiency with 50% fewer items than the PPT procedure. The overall results imply that the CAT procedure will be useful when an English placement test is administered at the beginning of each semester. Less
|
