The simulating study of the item exposure rate in computerized adaptive tests

• Main Authors: Yu Shin Hsieh, 謝友詩
• Other Authors: Hsiang Chuan Liu
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/26440347656695851377

碩士 === 臺中師範學院 === 數學教育學系 === 93 === For operational computerized adaptive tests, the most important issue is the overexposure item rates. The item having overexposure rate means most of the examinees tested it. When the examinees retest, they tend to test the same items, which leads to serious test security and equity risks. In this study, discuss the effects of the five item selection criterions — minimum offset difficulty, maximum interval information, KL information, NN criterion, and STR-B — were compared with respect to the precision of the trait estimation and the effect of the item usage at the same item banks. In the result, by the exposure rate and the precision the selection criterions could separate to three groups: maximum interval information and KL information criterions which having more precision of estimation; minimum offset difficulty criterion and STR-B which having more uniform exposure rates; NN criterion which balancing the estimation precision and effective item usage.