探討台灣青少年非肢體攻擊行為及受害經驗之雙向關係－隨機截距交叉延宕模型分析

Abstract

過去文獻關於非肢體攻擊與非肢體受害間的長期關係研究結果不一致。青少年可能會產生非肢體攻擊行為也可能遭受非肢體攻擊，更可能同時具有非肢體攻擊行為又有非肢體受害經驗。非肢體攻擊與非肢體受害孰先孰後仍有待研究。先前的研究未能將人與人之間的關係以及個人自己隨時間的改變區分開來。本研究運用隨機截距交叉延宕模型來釐清非肢體攻擊行為和非肢體受害經驗之間的長期關係。 本研究追蹤1,365名國高中學生四個學期，每學期調查一次。學生們在每所學校的電腦教室內填寫問卷，共有856名學生（449名女孩，407名男孩）提供了完整的四波數據。每波之間的間隔為六個月。 隨機截距交叉延宕模型的結果發現，非肢體攻擊行為和非肢體受害經驗的隨機截距間有正相關。非肢體攻擊行為和非肢體受害經驗的穩定性（自迴歸關係）顯著，代表青少年現在非肢體攻擊行為和非肢體受害經驗越高下一波也越高。值得注意的是，非肢體攻擊行為對非肢體受害經驗具有顯著的負向延宕效應，但是，非肢體受害經驗對非肢體攻擊行為沒有呈現延宕效應，這表示非肢體攻擊行為會降低下一時間點的非肢體受害經驗，但是非肢體受害經驗與下一波的非肢體攻擊行為沒有顯著相關。另外，兩者在四波均呈現顯著的同期相關。再者，以上的關係並未呈現性別差異。這兩者之間的特質相關因子，可能是造成非肢體攻擊行為與非肢體受害經驗高度相關的原因。本研究結果對青少年非肢體攻擊行為與非肢體受害經驗間的關係有更多的了解，將有助於制定更有效的預防和介入策略。

In previous studies, the temporal relationships between non-physical aggression (NPA) and non-physical victimization (NPV) have shown inconsistent results. It's unclear whether NPA precedes NPV or vice versa. Adolescents involved in non-physical aggression incidents may engage in NPA, experience NPV, or both. Previous research has failed to distinguish between interpersonal effects and intra-individual changes over time. This study aimed to clarify these relationships using a random intercept cross-lagged panel model (RI-CLPM). It is hypothesized that NPA and NPV will have bidirectional associations and that trait-like effects will be significant.The study included 1,365 students from 24 junior high schools and 8 senior high schools in Taiwan. These students were followed for four semesters, and 856 of them provided complete data across four waves, with a six-month interval between each wave.The results indicated a significant positive correlation between the random intercepts of NPA and NPV. The stability of NPA and NPV was significant. Surprisingly, NPA had a significant negative lagged effect on subsequent NPV, but NPV did not show a lagged effect on subsequent NPA. The results confirmed a unidirectional negative effect from NPA to NPV, meaning that non-physical aggression decreases the subsequent non-physical victimization, the study clarified the causal direction between NPA and NPV at both between-person and within-person levels. Trait-like effects may lead to becoming both a perpetrator and victim. Overall, this study, through the application of RI-CLPM, clarified the causal direction between non-physical aggression and non-physical victimization and found that trait-related factors significantly influence the high correlation between NPA and NPV. Understanding the relationship between NPA and NPV in adolescents can help develop more effective prevention measures.