《台灣社會學》

本研究運用一項長達三年，有關青少年成長與發展調適問題的長期貫時性追蹤資料（樣本數約為 1,400 位青少年），嘗試突破傳統青少年行為發展研究在方法上的困限，並試圖釐清研究文獻上有關「青少年憂鬱症狀與偏差行為併發之關係機制」的理論爭議。應用潛在成長曲線模型(latent growth curve model)分析，本研究發現：青少年的憂鬱症狀與偏差行為之間，並無心理或行為發展學者所強調之「潛藏在心理深層結構的因果聯結」，青少年憂鬱症狀與偏差行為併發的現象，主要應是由「外力」所促發。本研究並以實證資料說明傳統ARCL 模型在分析個人的成長或變化軌跡時，所遭遇的方法上之缺失，進而展示潛在成長曲線模型在研究「個人成長或變化」的個別差異時之優勢與適切性。
本研究同時發現：青少年男女之憂鬱症狀或偏差行為的發展，有相當不同的發展途徑，這可能是青少年男女經歷不同的社會化途徑所造成的結果。

Based on a panel study (sample size approximately 1,400) of adol-escent development and adaptation problems, the authors try to break through the limitations of traditional research methods used in studies of adolescent developmental trajectories. An attempt is made to clarify deba-tes regarding the co-occurrence mechanism of depressive symptoms and adolescent delinquency. Building on Latent Growth Curve Model analy-ses, we show that the "latent inner causal relationships" between depress-ive symptoms and delinquency (which many developmental psychologists and human development theorists enthusiastically embrace) cannot be sup-ported empirically. Instead, "outside forces" may best explain most co-oc-currences. We also discuss some of the disadvantages of the Auto Regres-sive Cross-Lagged Panel Model (ARCL) model in studying individual gro-wth or development and some of the advantages of the Latent Growth Cur-ve Model in analyzing individual developmental trajectories. It is noted that male and female adolescents follow significantly different paths when developing depressive symptoms and delinquent behaviors, perhaps best explained by their respective socialization processes.