教育测验中常出现使用一组相互联系或制约的项目，即题组。题组测验不服从项目反应理论的局部独立性假设。因此，此前研究者提出了基于项目反应理论变式的题组反应模型。然而，该模型对于两个因素还欠缺考虑：题组内项目的排列顺序和题组内项目间的特殊相关关系。

本研究基于某中学八年级的数学测试成绩，通过在题组模型中增加顺序参数，以解决题组内题目排列先后顺序的影响。题组的拟合中，纳入题组参数和同时纳入题组参数和顺序参数的模型均比传统项目反应模型效果要好。且后者的拟合效果略优于前者，说明了纳入顺序参数的必要性。

在参数估计的结果中，题组出现了题组参数方差不显著，却在顺序参数的方差上有显著结果的现象。这说明顺序参数能够补充解释题组中题组参数无法解释的部分。

A set of interrelated or restricted question groups often appears in educational tests. The testlet is contrary to the independence assumption of the item response theory. Therefore, researchers have previously proposed a testlet response model based on the variation of item response theory. However, the model still lacks consideration of two factors: the order of the items in the testlet and the special correlation between the items in the testlet.

The study is based on math test scores of the eighth grade of a middle school. By adding sequence parameter to the testlet response model, the influence of the order of the items in the testlet is covered.In the fitting of tesetlet, the models that incorporate the testlet parameter and include both the testlet parameter and the sequence parameter are better than the traditional item response model. And the fitting of the latter is slightly better than the former, which shows the necessity of including the sequence parameter.

In the result of parameter estimation, there is the result that the variance of the testlet parameter is not significant, but there is a significant result on the variance of the sequence parameter.It shows that the sequence parameter can supplement the part of the explanation of the testlet that the testlet parameter cannot explain.