本研究主要目的在於探討手足結構對於青少年偏差行為之影響與關聯性。在探討家庭與青少年偏差行為相關研究中，大部分學者都將焦點放在父母對於青少年偏差行為之討論，而忽略了手足對於青少年影響的重要性。因此本研究試圖與以往談論家庭對於青少年偏差行為的重心不同，而是將焦點放在手足對於青少年之影響，以期建構一個更加完整的架構來解釋家庭與青少年偏差行為之關聯性。本研究資料是以「台灣教育長期追蹤資料庫」（Taiwan Education Panel Survey，簡稱TEPS）2005年第三波調查之學生與家長問卷，研究對象為台灣地區94學年度就讀高中、高職、五專二年級之學生（N=12,643），並運用多元迴歸分析方式進行資料處理。

論文外文摘要： 

Previous studies have focused on how peers and family influence adolescent delinquency. Here, researchers have emphasized the importance of parental influences. Parents exert influence through both modeling and parenting. On the other hand, previous studies on peer influence have consistently demonstrated that youths who affiliate with deviant peers tend to adopt their delinquency through the social learning process. However, most of these studies have ignored the influence of other family members, especially siblings. This is a notable omission since siblings are the second closest members （next to parents） to youths in the family context. Siblings are also the first peers to these youths. Because previous studies have not investigated the importance of this intersection, this study explored the influence of siblings on adolescent delinquency. This research therefore proposed a sibling structure concept consisting of two version of conceptualization. The first version involves a delinquency comparison between only children and children with siblings. This demonstrates how being an only child affects adolescent delinquency. The second version involves a consideration of birth order when children have one or more siblings. Here, there are three categories （i.e., only child, child with older siblings, and child with younger siblings）. This study thus expanded on the traditional dichotomous view of the sibling effect to further our understanding of adolescent research. This study’s data were taken from the Taiwan Education Panel Study （TEPS） （a large panel study in Taiwan）. Specifically, the fourth-wave dataset was used because it contained sibling-related measurements. These data were collected from eleventh-grade students and their parents. The final analytic sample contained 12,643 adolescents after list-wise deletion. Multiple regression models were then used to test the effects of sibling structure. Analyses were separately conducted according to sex to explore any relevant differences in these effects. Results indicated no significant difference in adolescent delinquency between only children and those with siblings. However, the delinquency levels of non-first-born children were higher than those of first born and only children when birth order was considered. The sex-based analysis indicated that being an only child and first- born played a protective factor for males. In addition, having siblings resulted in more deviance than being an only child. In regard to birth order, the delinquency levels of older children were lower than those of non-first-born children. For female adolescents, there were different results. That is, being an only child had no effect on adolescent delinquency in this context. However, the delinquency levels of first-born adolescents were lower than those of non-first-born adolescents when birth order was considered. There was no significant difference between being an only child and being a non-fist-born child for female adolescents. Lastly, older sisters generally had higher levels of delinquency for female adolescents.