Title :
An Improving Technique of Adaptive Testing System Based on IRT
Author :
Wan, Hongxin ; Peng, Yun
Author_Institution :
Math. & Comput. Sci. Coll., Jiangxi Sci. & Technol. Normal Univ., Nanchang, China
Abstract :
Adaptive testing is a form of intelligent computer test. The number and types of the exam questions are uncertain. How to design high reliability algorithms to reflect the real level of the candidates is the key problem, also it is the most difficult problem. The realizing algorithms of parameter set, iterative computation and answer information estimation are designed and improved in the paper. The adaptive testing theory, the testing system design idea, the function and the technique are also introduced by an actual system designation. By comparing with the conventional test, the effect of the adaptive testing system is analyzed and evaluated.
Keywords :
computer testing; estimation theory; iterative methods; IRT; adaptive testing system; adaptive testing theory; answer information estimation; high reliability algorithms; intelligent computer test; iterative computation; realizing algorithms; system designation; testing system design idea; Adaptive systems; Algorithm design and analysis; Automatic testing; Computer science; Educational institutions; Gaussian distribution; Iterative algorithms; Mathematics; Parameter estimation; System testing;
Conference_Titel :
Biomedical Engineering and Computer Science (ICBECS), 2010 International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-5315-3
DOI :
10.1109/ICBECS.2010.5462368