1988 Fiscal Year Final Research Report Summary
Characteristics of the multiple choice-questions intrinsic to their format and their effect on the education
| Project/Area Number |
62510149
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Educaion
|
|---|
| Research Institution | KAWASAKI MEDICAL SCHOOL |
|---|
Principal Investigator |
SAITO Taiichi Kawasaki Medical School. Professor, 医学部, 教授 (00048258)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
NASU Ikuo Nihon University. School of Dentistry at Matsudo. Instructor, 松戸歯学部, 講師 (80112952)
ARITA Seizaburo Kawasaki Medical School. Associate professor, 医学部, 助教授 (20098601)
|
|---|
| Project Period (FY) |
1987 – 1988
|
| Keywords | MCQ / Random guess / Correct terminal / Key varidation / 車後評価 / 医師 ・歯科医師国家試験 / 共通一次試験 / 良い問題形式 |
|---|
| Research Abstract |
Multiple choice questions (MCQ) are not suitable for testing an examinee's knowledge. Examiner cannot distinguish whether questions have been answered by random guessing or not. In many types of MCQ, the knowledge of only two items (a correct terminal) out of four or five is sufficient to answer a question correctly even if all other items are blank. On the other hand, even if an examinee knows all three correct items appearing in the answer code, he is still compelled to make a random guess. Therefore, the scores achieved by the examinees on such tests can not be said to truly reflect their actural knowledge of a subject.
The correct answer exists always only one and examinees must choice only one answer code among presented four or five codes. Therefore, students try to find only one item among answer codes, and they lose their thinking ability, the freedom of thinking, studying habbit in the books and so on. Students become uniform and lose their individuality. Examiners try to evaluate the tests themselves, not their students, and apply many indices such as phy-coefficient and standard variation. We proposed the proper format of MCQ which can testify the real knowledge of examinees.
|
