中文摘要：

认知诊断模型-DINA模型假定属性之间是相互独立、非补偿的关系,但在实际应用中,属性之间有可能存在补偿关系。文章探讨了在不同的被试总体下,DINA模型在属性之间独立和存在补偿关系时作答数据上的表现,结果表明,当测验中有部分属性间存在补偿关系时,会给DINA模型诊断分布带来较大的影响,在固定其他条件（被试人数、随机因子）条件下,存在和不存在补偿作用所导致的诊断分类率之间有显著差异,且在不同大小补偿作用下,诊断分类率之间的差异显著;随着随机因子的增加,不同大小的补偿因子所导致的诊断分类率之间的差异会变小,补偿因子和随机因子之间存在着显著的交互作用。

英文摘要：

The cognitive diagnostic DINA model assumes that the attributes are independent and non-compensated, but possibly exists a compensation relationship between attributes in practice. This paper discusses the performance of the DINA model in the data of independent and compensating relationship between properties under different subjects. The study result shows that the compensation relationship if there is between some properties in the test will greatly affect the diagnosis distribution of the DINA model. There exists significant difference of diagnosis classification rate between the cases with compensation and without compensation effects under such fixed conditions as number of subjects, random factors. Moreover, the difference between the diagnostic classification rates is significant under different-magnitude compensation effects. With the increase of random factor, the differences of diagnosis classification rate caused by different-magnitude compensation factors become smaller, and there exists a significant interaction between the compensation factor and the random factor.