自尊、親子與師生關係對青少年至成年初顯期憂鬱症狀發展軌跡之探討

憂鬱症狀的發展是心理與教育界甚為重要的議題，但從青少年到成年初顯期的軌跡卻少被探討。本研究 藉由「臺灣青少年成長歷程研究」資料庫中2002至2009年間三波次(國三、高三、大四)共992位固定 樣本採用二因子潛在成長曲線模式之隨機截距隨機斜率模型，來探討自尊、親子與師生關係對憂鬱症狀 之發展軌跡。研究發現憂鬱症狀除穩定下降的一組外，另還有上升的一組。有以下四個重要結果: (1)自尊、親子與師生關係的國三分數皆負向影響後續變化幅度。另，高憂鬱組在此三變項的平均分 數皆較低憂鬱組低，且波動也較大。(2)自尊的重要性相對高於親子與師生關係;不論高低憂鬱組， 國三自尊分數皆負向影響憂鬱症狀分數，自尊的變化幅度也負向影響憂鬱症狀的變化幅度。但在高憂鬱 組，國三自尊分數卻對憂鬱症狀之變化幅度有正向影響，顯示過高自尊可能的問題。(3)低憂鬱組的 國 三親子關係負向影響憂鬱症狀分數，但高憂鬱組之親子關係變化幅度會負向影響憂鬱症狀變化幅度。 (4)最後，低高憂鬱兩組的師生關係分數對憂鬱症狀變化幅度皆無明顯影響，僅在低憂鬱組的國三師 生關係分數負向影響憂鬱症狀分數。本研究發現自尊、親子與師生關係在青春期到成年初顯期，會因憂 鬱症狀高低而對個體有不同效果，並據此提出教學與輔導相關建議。  Depression is an increasingly important area in psychology and education. Investigating depressive symptoms is a continuing concern for adolescents. Recently, an increased interest extends to emerging adulthood. According to the viewpoint of ecological system theory, previous studies in process-person-context-time model have highlighted the important factors such as ideal self, parent-child relationship, teacher-student relationship. This study aimed to investigate the trajectories of depressive symptoms from adolescence to emerging adulthood by considering three predictors — self-esteem, parent-child, and teacher-student relationships. Using AMOS 23.0, two factor latent growth curve model with random intercepts and random slopes was used to analyze the longitudinal data. In this panel study, a sample of 992 students were selected from the Taiwan Youth Project (TYP) in three waves from 2002 to 2009 (at Grade 9, Grade 12, and the senior year of undergraduate). The average scores in each of the three waves served as indicators of latent intercepts and slopes for each of the four study variables. By using hierarchical cluster analysis with dendrogram, we were able to divide all subjects into two groups as low and high depressive symptom groups respectively. Four research questions with correspondent eleven research hypotheses were investigated. That is, H1a: The intercept of self-esteem has negative influence on the correspondent slope of self-esteem. H1b: The intercept of parent-child relationship has negative influence on the correspondent slope of parent-child relationship. H1c: The intercept of teacher-student relationships has negative influence on the correspondent slope of teacher-student relationships. H2a: The intercept of self-esteem has negative influence on the intercept of depressive symptoms. H2b: The intercept of self-esteem has negative influence on the slope of depressive symptoms. H2c: The slope of self-esteem has negative influence on the slope of depressive symptoms. H3a: The intercept of parent-child relationship has negative influence on the intercept of depressive symptoms. H3b: The slope of parent-child relationship has negative influence on the slope of depressive symptoms. H4a: The intercept of teacher-student relationships has negative influence on the intercept of depressive symptoms. H4b: The intercept of teacher-student relationships has negative influence on the slope of depressive symptoms. H4c: The slope of teacher-student relationships has negative influence on the slope of depressive symptoms. The main findings of this study are shown as follows: (1) The intercepts of self-esteem, parent-child, and teacher-student relationships all negatively affected their respective slopes. In addition, the intercepts of the high depressive group on these three variables were lower, and the slopes were greater, than those of low depressive group. (2) Among the three predictors, self-esteem was the most important one. The intercept of self- esteem affected the intercept of depressive symptoms; the slope in self-esteem also affected the slope in depressive symptoms in both groups. Surprisingly, in the high depressive symptom group, people with high self-esteem in Grade 9 had a higher positive depressive symptom slope in the future. (3) The intercept of parent-child relationship negatively affected the intercept of depressive symptoms for the low depressive symptom group, while the slope in parent-child relationship affected the slope in depressive symptoms for the high depressive group. (4) The intercept of the teacher-student relationship negatively affected the intercept of depression symptoms only in the low depressive symptom group. Based on these results, suggestions for practical applications and future research are also discussed. In brief, it is necessary to pay attention to students’ development of long-term self-esteem and other psychological and social factors even if an individual has good self-esteem from the beginning. It is worth noting that individuals with high self-esteem are likely to have greater changes in depressive symptoms in the future. If an individual has high self-esteem but cannot tolerate setbacks and failures, the slope in depressive symptoms will increase. Moreover, even if individuals have unfavorable family experiences in the early stages, they still have the possibility of improving parent-child relationships, showing that the parent-child relationship is not limited to a single point in time. Finally, the teacher-student relationship in early adolescence not only shows its long-term influence, but also becomes an individual’s protective factor for depression, helping individual to have better adaptability in future.