本研究旨在探討國中班級中，導師班級經營、班級氣氛與班級關係攻擊情況之間的關係。本研究以台北縣市及彰化縣七所國中之 50 個班級為研究對象，共計 50位導師與 1761位學生參與研究。研究工具包括「導師班級經營風格量表」 、「班級氣氛量表」、「關係攻擊同儕提名問卷」、與「關係受害經驗量表」 。研究結果為：(1)單因子變異數分析結果發現，不同類型導師經營風格之班級關係攻擊情況沒有顯著差異。(2)導師「同情受害者」的態度與班級關係攻擊情況達顯著正相關；而根據迴歸分析的結果，導師對關係攻擊的評估與態度並無法預測班級關係攻擊、但能預測班級關係受害，且以「同情受害者」態度的預測力最大；導師對關係攻擊之處理策略也無法預測班級關係攻擊情況。(3)班級氣氛中的「同學支持」 、 「組織」 、 「團結」與班級關係攻擊達顯著負相關；班級關係受害與「同學支持」及「團結」有顯著的負相關、而與「班級關係攻擊信念」有顯著正相關；迴歸分析結果發現，班級氣氛與班級關係攻擊信念無法預測班級關係攻擊、但能預測班級關係受害，且以「班級關係攻擊信念」的預測力最大。最後，研究者依據研究結果對教育及未來研究提出具體建議。

The purpose of this study was to investigate the relationships among teachers' classroom management styles, classroom climate, and different levels of relational aggression and relational victimization. The sample consisted of 50 classes, 50 teachers and 1761 students from Taipei and Changhua. The instruments included Teachers' Classroom Management Style Scale, Classroom Climate Scale, Relational Aggression Questionnaire – Peer Nomination, and Victimization Experiences of Relational Aggression – Self Report Scale. The main findings were: 1. There were no significant differences in classroom relational aggression and victimization among the classes with teachers of different classroom management styles. 2. Teachers' sympathy for the victims was significantly related to the classroom relational aggression and relational victimization respectively. Teachers' appraisals and attitudes toward relational aggression could not predict the level of classroom relational aggression, but could predict the level of classroom relational victimization. Teachers' strategies for dealing with relational aggression could not predict the levels of classroom relational aggression and relational victimization. 3. There were significant correlations between student support, organization, cohesion and the level of classroom relational aggression; also there were significant correlations between student support, cohesion, relational aggression beliefs and the level of classroom relational victimization. The variables of classroom climate and relational aggression beliefs could not predict the level of classroom relational aggression, but could predict classroom relational victimization.