计算机自适应测验中RASCH模型稳健性的模拟研究
Simulation Study of Rasch Model’s Robustness in Computerized Adaptive test(CAT)
计算机自适应测验（Computerized Adaptive Test，以下简称CAT）是建立在项目反应理论（Item Response Theory）的基础上，按照一定的策略进行选题并根据被试的作答情况不断地调整测题，以准确地估计被试的能力水平。本研究采用模拟数据的方法，在CAT中分别采用Rasch及Birnbaum两种模型估计能力，通过比较两者的误差均方根（Root Mean Square Error，简称RMSE）、平均差异（Average Deviation，简称AD）及能力相关，对Rasch模型在CAT中的稳健性进行了研究。结果发现RASCH模型在区分度不等的条件下中仍然能较准确地估计被试的能力水平，具有很强的稳健性。
omputerized adaptive test (CAT) is based on Item response theory, select item according to some rule of Item-selecting and examinees’ responses to items to estimate the value of ability accurately The paper uses the method of simulating data, estimate the value of ability with Rasch model and Birnbaum model respectively in the CAT, study the Rasch model’s robustness through comparing Root mean square error(RMSE) and Average deviation(AD) and correlations of ability between Rasch model and Birnbaum model, the outcome finds the Rasch model in CAT is robust even if Rasch Model’s assumption (equal item discrimination parameters) can’t be satisfied.
邓远平、罗照盛
计算技术、计算机技术
，Rasch模型，Birnbaum模型，稳健性
RASCH Model Birnbaum Model Robustness
邓远平,罗照盛.计算机自适应测验中RASCH模型稳健性的模拟研究[EB/OL].(2005-12-22)[2025-08-16].http://www.paper.edu.cn/releasepaper/content/200512-554.点此复制
评论