同伴教学法（Peer Instruction 简称PI），PI教学方法是哈佛大学著名教授Eric Mazur创立，并且在世界许多国家正在使用的教学方法，PI教学法的一个关键环节是在课堂教学中基于概念测试题(ConcepTest)利用Clicker先让学生回答个人的答案，然后邻座讨论，讨论后再次发送答案，这个过程可以激发学生对概念问题的讨论、争辩和相互间的挑战，实现课堂上学生之间的合作学习和师生互动。因此，概念测试题的设计对成功地使用PI教学法有至关重要的作用。2010年起北京师范大学在基础物理电磁学课程中开始使用PI教学法，本研究的目的是，设计、编制和研究适合这个课程目标和学生水平的概念测试题，研究通过在使用PI教学法教授的电磁学课程中使用这些概念测试题，利用Clicker课堂交流系来收集学生答题的数据，利用SPSS等统计软件对收集的数据进行统计分析和研究，得到如下结果：1、PI教学法通过概念测试题和学生之间的讨论能够促进学生对概念的理解：讨论后94.6%的题目的正确率都有了很大的提升，所有的学生在讨论后正确率都要高于讨论前的正确率，讨论时学生彼此之间的正影响要大于负影响，概念测试题的难度在20%～70%之间讨论均有效；2、相对于男生来说，女生在讨论中获益相对更多；3、原本基础较好的学生从讨论中获益更多；4、对于异质固定分组与随机分组的学生，固定分组的同学讨论后的正确率、正影响均要略高出随机分组的同学，固定分组促进小组成员得到正确答案。

Peer Instruction (referred to as PI), PI teaching method is founded by Professor Eric Mazur from Harvard University, which is being used in many countries all around the world. The most important part of the PI is implemented in this way: firstly, a question which focuses on core concept named ConcepTest is given. Secondly, students send their personal answers to the teacher by clicker system. Thirdly, they discuss the question with each other in small groups which would stimulate debates and mutual challenges. Lastly, students send their answers again to the teacher. All these processes make the class be cooperative and interactive. So, the ConcepTest is very important for using PI successfully. We begin using Peer Instruction to teach basic physics course in Beijing Normal University since 2010. The purpose of this study is to design ConcepTest questions which matches the course objectives and suitable for the students of this level. Then the ConcepTest questions are used in class. Data are collected by clicker system and analyzed by SPSS and EXCEL. Results are as follows:1. Disscussion and ConcepTest can promote the students’ understanding of concepts. Not only 94.6% of the conceptests’ correct rate raised but also all the students’s correct rate became higher after discussion, the positive effect of the discussion is higher than the negative effect. ConcepTest of difficulty between 20% and 70% would make discussion effective.2. Girls benefit more from the discussion than boys. 3. Students with better foundation will benefit more from discussion. 4. The correct rates and conversion rates of fixed grouping students are slightly higher than the random grouping students. Team members would be promoted to get the right answer in fixed groups.