本研究採用臺灣教育長期追蹤資料庫的調查資料，探討中學生整體快樂及學校快樂的潛在成長模型與彼此之間的關係，並檢驗這兩種快樂與學業能力的關係。潛在成長模型及交叉延宕相關分析結果顯示，台灣中學生的整體快樂及學校快樂程度皆隨著年級而下降，但是兩者的初始水準愈高，下降幅度則較慢。其次，交叉延宕相關分析顯示整體快樂及學校快樂可相互預測。第三，學業能力初始水準愈高，成長速度愈快。第四，學校快樂初始水準與學業能力初始水準為顯著正相關，但學校快樂初始水準愈高，學業能力成長速度較慢。第五，學校快樂下降速度較快，學業能力成長速度也較快。最後，國一時的學校快樂正向預測國三學業能力，後者正向預測高二學校快樂。意味著在相同教育階段內，學業快樂正面影響之後的學業能力；在教育階段轉換之間，學業能力正面影響學校快樂。根據本文發現，學生快樂是心理學解釋與經濟學解釋交互作用下的結果。最後討論未來研究方向及可能的實務做法。

This research presented latent growth models of general happiness and school happiness among Taiwanese high school students. The relationships of these two types of happiness and academic competence were also examined. Seven hypotheses were made based on the economic and psychological explanations of happiness, the academic emotion theory, the broad-and-build theory, and the theory of eudaimonic well-being. The data from the Taiwan Educational Panel Survey were analyzed, and latent growth modeling and cross-lagged correlation were conducted. First, it was found that high school students showed decreasing general happiness and school happiness as grade level increased. However, there were slower declines for those with higher initial levels of general happiness. Second, the cross-lagged correlations showed that general happiness and school happiness significantly predicted each other. These findings implied that the economic explanation would better explain the origin of happiness than the psychological explanation. Third, the relationship between general happiness and academic competence was not significant. Fourth, the higher the initial level of school happiness, the slower the academic competence grew. The growth slope of academic competence was correlated positively with the growth slope of school happiness. But the growth slope of school happiness was negatire. Students with faster growth of academic competence would experience faster decline of school happiness. It implied that academic competency might grow at the cost of decreasing school happiness. These findings did not support the prediction from the academic emotion theory, broad-and-build theory, or theory of eudaimonic well-being. However, the theories of academic emotion and eudaimonic well-being were supported by results from the cross-lagged correlations. It showed that school happiness in Grade 7 positively predicted academic competency in Grade 9, which in turn predicted school happiness in Grade 11. It suggested that in the same educational stage, school happiness affected academic competence; while transitioning between two educational stages, academic competency affected the following school happiness. Further research and practical implications were discussed.