An Examination of the Literature on Criterion-Referenced and Computer-Assisted Testing
Gorde:
| Egile nagusia: | |
|---|---|
| Argitaratua: | |
| Gaiak: | |
| Sarrera elektronikoa: | Citation/Abstract Full text outside of ProQuest |
| Etiketak: |
Etiketarik gabe, Izan zaitez lehena erregistro honi etiketa jartzen!
|
| Laburpena: | Criterion-referenced testing (CRT) is defined as a method of ascertaining an individual's status with respect to some performance standard. Computer-assisted testing (CAT) is a method of constructing tests using a variety of computer techniques such as a single test computer printouts, stored item banks, teacher specified criteria, machine readable answer sheets, etc. After an examination of the literature on both subjects, the conclusion reached is that CRT and CAT may help each other in the following ways: (1) item generation techniques may be refined to allow more comprehensive evaluation of domains by making more items available; (2) item sampling algorithms may be used to achieve more representative tests from existing domains; (3) branching tests may be utilized to arrive at the most cost-effective method for evaluating performance; (4) test models may be simulated to ascertain their feasibility; (5) mathematical models may be developed to help define and standardize the criteria by which performance is judged; and (6) CRT can be used more widely as a valid theory to aid the design of CAT systems. There is a 16 page annotated bibliography divided into two separate subject divisions. (Author/NR) |
|---|---|
| Baliabidea: | ERIC |