偏遠地區與一般地區國中生學習成就差異比較：以「反事實」分析

A Comparison of Differences between Learning Achievements of Students in Remote Areas and Students in General Areas: A Counterfactual Analysis

本研究以「反事實」分析，應用傾向分數配對法（Propensity Score Matching， PSM）估算平均處理效果值（Average Treatment Effect on the Treated， ATT），進行 偏遠地區與一般地區國中生綜合分析能力學習成就的差異比較。本研究的分析資料 採用「台灣教育長期追蹤資料庫」（Taiwan Education Panel Survey， TEPS）之2001 年第一波（七年級）和2003 年第二波（九年級）的國中追蹤樣本資料，共計有效 樣本數16，687 人。研究結果發現：(1)在還沒有控制個體基準線差異的條件情況之 下，偏遠地區國中學生的綜合分析能力學習成就平均數明顯低於一般地區，偏遠地 區國中生與一般地區國中生在學習成就有明顯的差異現象；(2)當控制了個體基準 線的差異性之後，以學習成就整體平均效果值來看，相較於一般地區，偏遠地區國 中生綜合分析能力學習成就並無太大落差。文末，根據研究結果進行討論。

In this study， propensity score matching with counterfactual analysis were used to estimate the average treatment effects of studying in junior high schools on learning achievements. Specifically， this study compared the differences between the comprehensive analytical ability of junior high school students in rural areas and general areas. We used data from 16，687 junior high school students from the first wave of 2001 (7th grade) and the second wave of 2003 (9th grade) of the Taiwan Education Panel Survey Data set. The results revealed the following significant findings: (1) When the differences between individual baselines were not controlled， the comprehensive analytical ability of students in rural areas are lower than that of students in general areas. This indicates that the comprehensive analytical ability of learning achievements from junior high school students in rural and general areas differ. (2) When the differences between individual baselines were controlled， the comprehensive analytical ability of learning achievements of students in rural areas are not lower than that of students in general areas. Based on the findings， the implications for research and practice are discussed.