The Effects of Higher-order item response theory

• Main Authors: Chia-Hua Lin, 林佳樺
• Other Authors: Bor-Chen Kuo
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/38867938059380820903

碩士 === 國立臺中教育大學 === 教育測驗統計研究所 === 97 === The assessment framework of the many large-scale standardized tests, such as the programme for international student assessment (PISA), is the higher-order assessment framework. The unidimensional item response theories (UIRT) are often used to estimate the overall ability and the multidimensional item response theories (MIRT) are often used to estimate the subscales. This estimation procedures is name as a separated estimation. However, there is no research on the effects of using the full models to estimate the overall ability and subscales concurrently. The main purpose of this study is to propose the higher-order IRT models being suitable for higher-order assessment frameworks based on the PISA and to estimate the subscales and overall ability concurrently. By using the simulation data, the performances of the two estimation procedures, the separated estimation used by PISA and the full model estimation proposed by this study, are compared. The root mean square errors (RMSEs) are the indicators of the performances. The results show that the performances of the full models proposed by this study are better than the PISA.