| Project/Area Number |
04680282
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
科学教育(含教育工学)
|
|---|
| Research Institution | Tokyo Institute of Technology |
|---|
Principal Investigator |
SHIGEASU Kazuo Tokyo Institute of Technlogy, Faculty of Enginering, Professor, 工学部, 教授 (90091701)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
NAKAMURA Tomoyasu Tokyo Institute of Technology, Faculty of Engineering, Instructor, 工学部, 助手 (30251614)
MATSUDA Toshiki Tokyo Institute of Technology, Faculty of Engineering, Associate Professor, 工学部, 助教授 (60173845)
ICHIKAWA Masanori Tokyo University of Foreign Studies, Faculty of Foreign Studies, Associate Profe, 外国語学部, 助教授 (20168313)
柴山 直 東京工業大学, 工学部, 助手 (70240752)
|
|---|
| Project Period (FY) |
1992 – 1993
|
| Project Status |
Completed (Fiscal Year 1993)
|
| Budget Amount *help |
¥2,200,000 (Direct Cost: ¥2,200,000)
Fiscal Year 1993: ¥800,000 (Direct Cost: ¥800,000)
Fiscal Year 1992: ¥1,400,000 (Direct Cost: ¥1,400,000)
|
|---|
| Keywords | Item Response Theory / Baysian Network / Gibbs Sampler / Baysan Statistics / CAI / 項目応答モデル / 学習者モデル / ベイズネットワーク / 自信度データ / 確率ネットワーク / マルコフ過程 |
|---|
| Research Abstract |
1) Estimatin of Competence and Confidence level Based on correct/wrong data as wel as confident/non-confident data, we deveoped the new IRT method which describes learners' positin in two dimensions, namely, true competence and confidence level.
2) Estimation of Knowledge State Combining traditinal IRT method and latent class model, we developed the method which classifies each learner into one of the thre casses, namely, "Compete Knowledge", "Partial Knowledge", and "complete Ignorance".
3) Estimationof posterior Distribution for Competence Parameter Using Gibbs Sampler, we developed a new method which derives the marginal posterior distribution for the competence parameters. When this distribution is available, we can make use of the formal decision theory to determine the next instruction step.
4) Baysian Network Representation of Leamers Based on the network model of problem solving process, we developed the method, which identifies each learner's state in this network, using the Baysian approach.
These methods provide useful information to design the ptimal sequence of instruction for eachlearner. Our goal is to construct CAI system which includes the adove mentioned methods to identify each learner's trace.
|
