本研究研究目的希望瞭解家庭因素與自殺危險性之間的關係是否為透過負向特質為中介變項的間接影響路徑，抑或是直接影響與間接影響路徑俱備。依據過去的研究，取自尊與問題解決能力為個體負向特質向度，以及與自尊、問題解決能力有關的家庭因素：家庭僵化、父母管教態度（嚴格，高期望、過度保護）與溝通為家庭因素的向度，以自殺危險性為探討的依變項，進行線性結構關係模式分析（LISREL）。 本研究以台北市信義區某國中一、二年級共411位男女學生為研究對象，有效樣本中一年級學生有190位，二年級學生221位，男生212人，女生199人，依研究變項選取已編製好且符合的量表為測量工具：包括父母管教態度量表、親子關係量表、親子溝通量表組合成「家庭分量表（一）（二）」，家庭適應與凝聚力量表為「家庭特性分量表」，自尊量表為「自我特質分量表（一）」，問題解決量表為「自我特質分量表（二）」，自殺危險程度量表為「生命興趣探索量表」。資料收集後以LISREL7為統計工具進行分析。 本研究的結果顯示家庭因素對自殺危險性而言，包括了直接影響與透過負向特質的間接影響，且家庭因素對自殺危險性的解釋力與預測力比負向特質大，顯示直接影響的效果比間接影響的效果要來得大。 此外，研究結果亦顯示將家庭因素不同向度獨立出來成為獨立的變項，來探討其與負向特質、自殺危險性之間的關係，可以得到對自殺危險性而言較好的解釋模式。顯示不同家庭因素的向度對自殺危險性有不同的影響方式，其中家庭僵化透過負向特質對自殺危險性有間接影響的效果，而父母管教態度的嚴格與高期待，對自殺危險性有直接的影響力。進一步探討父親與母親不同的管教態度與溝通對於青少年自殺危險性之間的關係為何，結果顯示父親過度保護透過負向特質，間接影響個體自殺危險性，而母親嚴格與期待的管教態度對個體而言，可以直接影響自殺危險性，也可透過負向特質間接影響自殺危險性。 / The purpose of this study was to detect the relationships of family factors, negative traits and suicidal risk and if the negative traits were the mediators between family factors and suicidal risk. This study predicted family factors indirectly affect suicidal risk. If not, the family factors would have both direct and indirect influences on suicidal risk. According to the previous researches, researcher took the family rigidity and child-rearing attitudes as the family factors, and the self-esteem and problem solving skills as the negative traits. The dependent variable was suicidal risk. The study applied the Linear Structure Relation（LISREL）to test its fitness. The 411 subjects were the junior high school students in Hsin-I area in Taipei City, including 212 boys and 199 girls. The measuring instruments were the Child-Rearing Attitudes Scale, the Family Adaptability and Cohesion Evaluation Scale II, the Rosenberg Self-Esteem Scale, the Personal Problem-Solving Scale, and the Suicidal Risk Scale. The data was analyzed by LISREL 7. The results revealed that the family factors affected suicidal risk both directly and indirectly. In terms of suicidal risk, the family factors were more predictive than negative traits. Besides, the present study also found that it would be a better model if taking the different dimensions of family factors as individual variables. And the different variables affected suicidal risk in different ways. The family rigidity indirectly affected suicidal risk through negative trait, and the sternness and high expectation attitudes of parents directly affected suicidal risk. Moreover there were differences between paternal and maternal child-rearing attitudes that affected suicidal risk. Paternal overprotection indirectly affected suicidal risk through negative trait. The sternness and high expectation of maternal attitudes affected suicidal risk both directly and indirectly.