近期校园欺凌在我国引起广泛关注。研究欺凌行为的产生机制，有助于从根本上来防治校园欺凌的发生。校园欺凌所发生的群体中，职业中学与普通中学经过中考的筛选，在同伴环境、学习氛围、生涯规划等方面都有所不同，社会媒体关于职业中学校园欺凌的报道层出不穷。因此，本研究以职业中学生为研究对象，探究校园欺凌的成因。 以下是本研究的理论依据、研究方法和研究结果： 理论依据：一般紧张理论解释了“人为什么要犯罪”的问题，社会控制理论解释了“人为什么不犯罪”的问题，校园欺凌与犯罪行为存在共性，也有所区别。本研究从这两个理论出发，对校园欺凌的产生机制进行分析和假设，建立研究框架，以期解决“一些学生为什么要欺凌别人”、“另外一些学生为什么不欺凌别人”的问题。 研究方法：通过问卷调查的方式对珠海市6所职业中学共1226名学生进行调查，实得有效问卷871份。采用描述性统计分析考察职业中学生欺凌与被欺凌的状况；采用独立样本T检验和方差分析考察校园欺凌状况在人口学变量上是否存在差异；最后在相关分析的基础上，建立校园欺凌成因的结构方程模型，探讨紧张和社会纽带对校园欺凌的影响，并检验消极情绪在其间的中介效应。 研究表明： 1.描述性统计分析结果显示，职业中学中，有超过三分之一的学生曾欺凌过其他同学，有超过一半的学生曾遭遇过同学欺凌； 2.T检验和方差分析结果显示，校园欺凌在性别和年级上具有差异性，遭遇欺凌的状况在父母受教育程度上具有有差异性； 3.结构方程模型结果显示，得而复失型紧张和挥而不去型紧张可以直接对欺凌行为产生显著正向作用，三种紧张类型都可以通过中介变量压力和倦怠间接地对欺凌行为产生显著正向作用； 结构方程模型结果显示，参与纽带和追寻纽带可以直接对欺凌行为产生显著负向作用，并都可以通过中介变量倦怠间接地对欺凌行为产生显著负向作用，追寻纽带还可以通过中介变量压力间接对欺凌行为产生显著负向影响。

The recent frequent occurrence of bullying on campus has aroused widespread concerns. The study on the generation mechanism of bullying behaviors is conducive to the prevention of school bullying radically. From the groups where school bullying happen, based on the screening and selection from vocational and middle schools, there are many differences in peer environment, learning atmosphere, career planning and so on between vocational middle schools and ordinary middle school. And the bullying behaviors reported by social medias in vocational middle schools emerge in endlessly. Therefore, this research takes vocational middle school as the research object to explore the causes of school bullying. Theoretical foundation, methods and results are as follows: Theoretical Conception: general tension theory explains the problem of "why people want to commit crimes", while social control theory explains "why people don't commit crimes", and school bullying and criminal behavior have some similarities and differences. From these two theories, this study analyzes and hypothesizes the mechanism of school bullying from the theoretical level, and establishes a theoretical framework. In order to solve the problem of "why some students want to bully others" and "why some students don't bully others", it can provide reliable theoretical support for empirical research. Methods: It surveyed 1226 students in 6 vocational schools in Zhuhai by means of questionnaires, and achieved 871 valid questionnaires. It uses descriptive statistical analysis to investigate the situation of bullying in vocational middle school students. Independent -samples T test and variance analysis are employed to analyze whether the school bullying situation has the difference in the demographic variable; Finally, on the basis of correlation analysis, it establishes the structural equation model of school bullying, Explores the influence of intension and social belt on school bullying and testify the intermedia effect posed by negative emotion. Research shows that: 1.In the light of descriptive statistical analysis results, more than 1/3 students have bullied other students and more than half of students have been bullied by their classmates. 2.Independent-samples T test and variance analysis indicates that, School bullying is different in sex and grade, and the situation of being bullied is different in education level of parents. 3.The structural equation model suggests that regained strain and sustain strain can have a significant positive effect on bullying directly, and all three types of strain can influence the bullying in an indirect manner through intermedia variable pressure and sluggish. 4.As the structural equation model showcases, engagement belt and pursuit belt pose a conspicuous negative effect on bullying through intermedia variable sluggish indirectly, and pursuit belt can pursuit belt pose a conspicuous negative effect on bullying through intermedia variable pressure indirectly.