项目反应理论(Item ResponseTheory,简称IRT)是一种新兴的心理与教育测试理论。项目反应理论是在分析与克服经典测试理论的局限性的基础上发展起来的，是在批评了经典测试理论的局限性的基础上发展起来的一种现代测试理论。项目反应理论比较成功的应用了实验误差控制的统计调整技术，在测验的较为微观领域即测验项目上开展研究，建立起项目反应模型，将测试项目“性质”对测量的影响参数化，再通过模型控制这些参数，从而达到控制测量误差的目的。项目反应理论的兴起和发展受到了测量理论和实践工作者们的高度重视，对心理与教育测量的发展起到了很大的促进作用。本文首先介绍项目反应理论的基本概念，模型，经典的模型参数的估计方法，然后对这些估计方法作实践，提出改进方案，最后把所有的算法作归纳、比较和整合，在软件中统一实现出来。

Item Response Theory (IRT)is a new psychological and educationaltesting theory. This theory is developed by analysis and overcomethe limited of traditional testing theory. It criticizes theweakness of traditional testing theory.IRT makes good use ofstatistical adjustment technology in difference control.It does theresearch on micro domain , that is on testing domain.It sets upproject model, parameterize the influence of the test item to thetest result, then control this parameter by the model, final resultin controlling the test inaccuracy.The rise and development of ItemResponse Theory is highly appreciated by the testing theory andexercisers, it pays a great promotion to the development ofpsychological and educational testing. This article mainly instructsthe base knowledge of Item Response Theory, the model of it, theestimation theory of the parameters of it, putting it into practiseby programme and the improvement of it in coding.