Erik Erikson（1968）認為自我認同為青少年最重要的發展任務。過去研究亦指出個體若越能有自我認同感，則會擁有正向的心理健康，及最適的心理功能表現（例如：自尊、幸福感）。此外，自尊亦為青少年重要的發展議題之一，它反映著個體的自我價值、自我形象及對自己整體的評價。過去諸多研究顯示，青少年的自尊問題會影響青少年的心理健康或心理病理（例如：憂鬱、行為偏差）。雖過去探討青少年自尊有諸多取向，但鮮少從自我認同來探討青少年的自尊。基於Erikson自我認同感之觀點，本研究認為青少年認同確定性越高，其自尊亦會越高。此外，由於不同青少年時期，其內在需求有所不同，因此本研究假設早、中期之青少年，其社會及形象認同重要性能預測其自尊，而晚期青少年則個人認同重要性能預測其自尊。本系列研究以逐步回歸分析探討不同階段青少年，其認同確定性及認同重要性是否能預測其自尊。研究一（共計1,285名國中、高中、大學樣本）支持研究假設，亦即認同確定性可預測青少年自尊，而認同重要性則隨不同的發展階段，對自尊有不同的預測效果。研究二（共計203名三所大學生樣本）成功複製研究一之結果。研究三（185名大學生樣本）及研究四（146名大學生樣本）更新認同確定性、認同重要性測量工具，得到與研究一、研究二相同的相同結果。總結，本研究建議認同確定性及認同重要性皆為青少年自尊重要的預測因子。

Adolescence is often regarded as a transitional stage between childhood and (emerging) adulthood. Adolescence also is viewed as a stage with heightened risks to a healthy development (Adams, Gullotta, & Markstrom-Adams, 1994; Roberts, Lewinsohn, & Seeley, 1996). Erik Erikson (1968) postulated that identity formation is the most important developmental task during adolescence. Researchers also suggested that a more confirmed sense of identity is more likely to lead to positive mental health (Erikson & Erikson, 1950; Marcia, 2002) and optimal psychological functioning (e.g., self-esteem; well-being) (Adler, Lodi-Smith, Philippe, & Houle, 2016; Lillevoll, Kroger, & Martinussen, 2013; Schwartz et al., 2011; Waterman, 1992). Besides, self-esteem is also an important developmental issue in the field of adolescence, reflected to as self-worth or self-image, is the global evaluative dimension of self (Harter, 2006). Researchers have been interested in self-esteem changes during adolescence in the long run (e.g., Harter, 2006; Kling, Hyde, Showers, & Buswell, 1999). For example, researchers generally found that self-esteem often decreases when children make the transition from elementary school to junior high school (Hawkins & Berndt, 1985; Twenge & Campbell, 2001). Robins and his colleagues also indicated that this decrease in self-esteem might take place during life transitions (e.g., from elementary school to junior high school, from junior high school to high school; from high school to college) (Robins, Trzesniewski, Tracy, Gosling, & Potter, 2002). In the past, several studies demonstrated that low self-esteem can influence adolescent mental health and psychopathology (e.g., depression, eating disorders, delinquency…) (Kuhlberg, Peña, & Zayas, 2010; Harter, 1993; Unger, Kipke, Simon, Montgomery, & Johnson, 1997; Usher, Zahn-Waxler, Finch, & Gunlicks, 2000). Numerous studies investigated adolescent self-esteem in various ways (e.g., Baldwin & Hoffman, 2002; Prinstein & Dodge, 2010; Roustit, Campoy, Chaix, & Chauvin, 2010), but the role of self-identity in self-esteem changes during adolescence has relatively received little attention. Based on Erikson's belief that a more confirmed sense of identity is more likely to lead to positive psychological adjustment (Erikson & Erikson, 1950) and help individuals ensure the direction, purpose, and meaning of life, this study proposed that all aspects of identity firmness are key predictors to self-esteem at all substages of adolescence. In addition, as different stages highlight different needs for the various aspects of identity, this study hypothesized that, in early and middle adolescence, social and image identity importance would significantly predict self-esteem; while in late adolescence, only personal identity importance would be significantly related to self-esteem. To examine the above hypotheses, four stepwise regression analyses of four serial studies on self-esteem using the aspects of identity importance and firmness as predictors were conducted in three different age levels of adolescents. Study 1 (N = 1,285 participants from junior high, high school, college students) generally supported the hypothesis that identity firmness could predict adolescent self-esteem. That is, the results of Study 1 showed that there were different patterns on identity importance in predicting self-esteem among three adolescent stages, but same patterns on identity firmness (i.e., personal identity firmness, social identity firmness) in predicting self-esteem among three adolescent stages. Study 2 (N = 203 participants from three different colleges) successfully replicated the findings of Study 1. In Study 3 (N = 185 undergraduate students) and Study 4 (N = 146 undergraduate students), the measurement of identity importance and identity firmness was revised. Again, the findings were consistent with those of the previous two studies. In sum, this study suggested identity importance and identity firmness are crucial factors contributing to adolescent self-esteem.