補習對數學成就馬太效應的影響：傾向值成長混合模型分析

The Effects of Cram Schooling on Math Achievement: A Propensity Score Growth Mixture Model Analysis

本研究主要探討補習對數學成就馬太效應的影響。本研究使用屏東教育長期追蹤資料庫的國小4年級至6年級學生固定追蹤樣本（N=9,739），並採用傾向值成長混合模型進行補習影響數學成就馬太效應的分析。研究發現補習的傾向值達到共變項平衡，而數學成就可潛在的區分成2個異質群體，且隨著時間的遞移，呈現強者恆強且弱者恆弱的馬太效應現象。而有補習的學生相較於沒有補習的學生，出現在強者恆強組相對於弱者恆弱組的機率較高。針對上述結果，提出相關的建議。

This study examines the impact of cram schooling on math achievement, specifically the Matthew effect. The study utilizes a longitudinal framework and analyzes data from the Pingtung Education Longitudinal Survey database using a propensity score growth mixture model. The study discovered that the propensity value of cram schooling reached equilibrium, and math achievement could potentially be divided into two heterogeneous groups. Over time, it exhibited the Matthew effect phenomenon, with tutored students having a higher probability of appearing in the higher grade group. Based on these findings, relevant suggestions are proposed.