本研究之目的在瞭解學生學業成就的成長情形及其與相關影響因素間的關係，因此利用「台灣教育長期追蹤資料庫」（TEPS）資料庫的四波調查，以二層次的階層線性模式（HLM）分析2,764位國中生的一般分析能力與數學分析能力的初始狀態及成長曲率，並探究其與隨時間變動的共變項間之關聯。研究發現，學生在國一時的一般分析能力及數學分析能力存在個體間的差異，且呈現顯著的先上後下之曲線成長。而一般分析能力及數學分析能力成長速率的變異，分別有59.61%及67.01%來自於時間變項的解釋；同時，本研究所討論的層次二變項所能削減的成長曲率變異數比例中，一般分析能力為25%，但數學分析能力僅有3%。再者，除時間外，隨時間變動共變項中的快樂程度對一般分析能力有影響，而期望─能力信念對數學分析能力有影響。最後，男生的一般分析能力及數學分析能力的初始狀態及成長曲率都較女生高，而家中子女數愈多，一般分析能力及數學分析能力的初始狀態也較差，但僅有數學分析能力的成長曲率受到家中子女數的影響。

Based on the four surveys of the TEPS database, this study applied two-level HLM analysis to studying the initial status and growth curvature of general analytic ability and mathematical analytic ability of 2,764 junior high school students. It also analyzed the correlation among covariates changed with time. The findings firstly showed that interindividual differences did exist in the general analytic ability and the mathematical analytic ability among first-year junior high school students and that growth curvature clearly went up and then went down. Furthermore, the variations of growth rate for general and mathematical analytic abilities could be attributed to time variables by 59.61% and 67.01%, respectively. Meanwhile, the variance proportions of the reduced growth curvature by level-two variables presented 25% for general analytic ability but merely 3% for mathematical analytic ability. Moreover, in addition to time, happiness in level-one variables was found to have effects on general analytic ability while expectation-competence belief was found to have effects on mathematical analytic ability. Finally, girls presented higher initial status and growth curvature in both general and mathematical analytic abilities than boys did. Besides, the initial status in general and mathematical analytic abilities was found to be worse in families with more children while only the growth curvature in mathematical analytic ability was found to be affected by the number of children in the family.