论文部分内容阅读
目的:探讨个人不拟合对IRT二参数模型项目参数估计的影响,并使用数据净化方法降低这种影响,提高个人拟合指标探测率。方法:基于二参数模型和lz指标进行分析。使用ICC面积法比较项目参数估计的变化,并使用数据净化的方法提高lz指标探测效果。结果:①不拟合被试比率越大,项目参数估计偏差越大;②增加测验长度可以降低个人不拟合对项目参数估计的影响;③加大样本量对降低个人不拟合对项目参数估计的影响没有作用;④数据净化方法可以有效的提高lz指标的探测效果。结论:个人不拟合会影响二参数模型的项目参数估计,数据净化方法可以校准项目参数估计,提高lz指标探测效果。
OBJECTIVE: To investigate the effect of personal mismatch on the parameter estimation of two-parameter IRT model and to reduce the effect by using data purification method and to improve the detection rate of personal fitting index. Methods: Based on the two-parameter model and the lz index. The ICC area method is used to compare the estimated changes in the project parameters and the data purification method is used to improve the detection performance of the lz indicator. Results: (1) The greater the proportion of non-fit subjects, the greater the deviation of project parameters estimation; (2) The longer the length of test can reduce the impact of individual non-fitting on the estimation of project parameters; (3) The estimated impact has no effect; ④ data purification method can effectively improve the lz indicator detection effect. Conclusion: Individual mismatch will affect the project parameter estimation of the two-parameter model. The data purification method can calibrate the project parameter estimation and improve the detection effect of lz index.