過去對於教育分流所做的研究很多，不過所使用的樣本大多距今已十幾年，而且年齡往往包括20-64 歲民眾，其中有許多年紀較大的樣本是在早年接受高中階段的分流教育。本研究主要根據台灣教育長期追蹤資料庫（簡稱TEPS）第三波2005 年之學生與家長資料做路徑分析，探討近年學生出身背景對高中階段教育分流影響所涉及的機會不均等性。研究結果顯示：雖然國中學生整體升學率已接近100％，不同出身背景者幾乎都能升學而無多少差異，可是影響所升上高中階段教育分流中，國中學業成績最具關鍵。家庭社經地位越佳、兄弟姊妹數越少，其國中學業成績較高者，有較高的比例升上出路較佳之公立高中。這充分顯示出量的機會不均等性雖趨近於0，但質的機會不均等性仍維持著，表示「教育機會不均等最大維持」（MMI）的假設未獲得支持，而支持「教育機會不均等（EMI）有效維持」的假設。

There were numerous previous studies on educational tracking. However, most of the data adopted in these studies were collected a decade ago, involving sample population 20-64 years of age. Among them, many elderly samples had undergone senior high school tracking in the early years. The present study used the data of students and parents from the 3rd wave (2005) of the Taiwan Education Panel Survey (TEPS). A path analysis was conducted to investigate the inequality of opportunities involved in the influence of student family background on educational tracking at the senior high school stage. The results show: although the overall percentage of junior high school graduates entering advanced levels was approximately 100% and almost all students of different family backgrounds entered advanced levels, the most crucial factor among those affecting educational tracking of senior high school stage was students academic performance during junior high school. Students with higher family socioeconomic status and fewer siblings showed better academic performance in junior high schools, and consequently had higher possibilities of entering public senior high schools leading to a more promising future. This indicates that though the inequality in the quantity of opportunities approached 0, the inequality in the quality of opportunities still remained, which fails to support the hypothesis of the maximally maintained inequality of educational opportunity (MMI); however, the result supports the hypothesis of the effectively maintained inequality of educational opportunity (EMI).