Adaptive computer testing based on the gradient methods

DOI №______

Abstract

The article deals with the gradient methods in adaptive computer testing for assessing the level of knowledge in the course of distant learning. It describes the possibility of increasing the accuracy of evaluation by determining the level of complexity of test questions in accordance with the current determined level of knowledge as well as ensuring the motivation of the test takers. The methodological approach to adaptive computer testing is described and the corresponding algorithm is suggested. The results of evaluation of the expected increase of testing efficiency by using adaptive computer testing on the example of development of a training course for the system of distant learning are analyzed.

Keywords: adaptive computer testing; distant learning; assessment of the acquired knowledge level; gradient methods; effectiveness of the testing algorithm.

References
1. Ositrov V. V. Terms and educational foundations control of troops of foreign language using modern technology of computer adapted testing [Umovy ta dydaktychni osnovy kontroliu rivni avolodinnia inozemnoiu movoiu viiskovosluzhbovtsiv z vykorystanniam suchasnoi tekhnolohii kompiuterno adaptovanoh otestuvannia] // Visnyk NAOU. 2009. No. 2(10). p. 56–62.
2. Fedoruk P. I. Adaptive tests: general provisions [Adaptyvni testy: zahalni polozhennia] // Mathematical Machines and Systems. 2008. No. 1. p. 115–127.
3. Fedoruk P. I. Using adaptive tests in intelligent control systems knowledge [Vykorystannia adaptyvnykh testiv v intelektualnykh systemakh kontroliu znan] // Shtuchnyi intelekt. 2008. No. 3. p. 380–387.
4. STANAG 600. [Electronic resource].— http://www.natobilc.org/files/file/6001EFed05.pdf.