本文以分量迴歸模型，運用「台灣敎育長期追蹤資料庫」探究親子教育跨代移轉之假說，並研究父母親教育以及親代教育程度差距，對於子女在校學習能力測驗分數分配變動影響。分量迴歸研究不僅顯示在同一條分數分配上，父母親教育差距衝擊不同分量上子女學習之成效，並可影響子女分數分配之高低差異，這是本研究特色。實證結果顯示，父母親敎育對子女在校學習成就具有顯著正向的影響，證實親子教育跨代移轉機制的存在。學歷越高的父親對於國中子女成績影響越大，子女成績分配之分量位置就越高。父親教育差異造成子女分數分配之差距，國中子女分數分配在低分量上差距大，在高分量上差距小，但擴大了高中子女在高分量上的成績差異。高中子女成績受到父親學歷影響較大，國中子女成績則受到母親學歷影響較高。研究所及以上學歷母親，子女在高分量上成績表現優異。比較相同教育父親和母親，母親對高、國中子女數學成績在高分量上的影響遠超過父親。

Utilizing more efficient estimators from quantile regression models and the reactions in quantile of the whole distribution, we analyze the systematic differences in intergenerational education correlations across the conditional quantiles of pupils' test scores. The key question is whether the bottom of the test score distribution looks different than the top. Our empirical evidence indicates a strong intergenerational transmission of educational inequalities between parental education levels and junior and senior school pupils' learning achievements. Parents with higher education affects pupil's learning in significantly positive ways. The differentials in senior and junior high school pupils' score distribution are positively associated to their father's education. Although these differentials persist in the high tail of these pupils' score distribution, there is convergence in the high tail of junior high school pupils' score distribution. The mother's education has a significantly higher influence on the learning distribution of junior high school pupils' score. By contrast, a father's education has a stronger impact on the learning distribution of senior high counterparts. Given a father, with the same education, a mother significantly and positively affects both junior and senior high school pupils' mathematics scores in the high end of the distribution.