家庭背景及選校策略對國中基測成績的影響

Abstract

本文使用臺灣教育長期追蹤資料庫(TEPS)2001及2003年資料與2004年基測成績，以OLS迴歸分析，分別從(1)學生基測原始成績、(2)學校層級的平均PR值變化、(3)學生個人層級的PR值變化，探討家庭背景及選校策略對國中基測表現與其相對PR值的影響。

透過實證分析顯示，學生早期能力(TEPS國一上綜合能力PR值)對基測PR值具有極高預測力，父母教育程度、家庭所得、就讀私立學校及都市地區對基測成績亦有顯著正向影響，家庭經濟變故則產生負面影響，學生國三階段的自習時間、校內外輔導與補習資源投入則顯著提升基測成績。

公私立與城鄉學校對基測表現的影響，透過控制其他背景與投入變數後明顯縮減，僅能部分解釋成績差異，顯示影響學生基測表現的更深層因素在於學生家庭背景、國中之前的早期能力與國三學習投入。

This study uses data from the 2001 and 2003 waves of the Taiwan Education Panel Survey (TEPS), along with Basic Competence Test (BCT) scores from 2004, to examine the effects of family background and school selection strategies on junior high school students’ academic performance and relative percentile ranks (PR). Ordinary Least Squares (OLS) regression analyses are conducted across three levels: (1) students’ raw BCT scores, (2) school-level changes in average PR values, and (3) individual-level changes in student PR values.

Empirical results reveal that students’ early academic ability—as measured by their TEPS first-year composite PR scores—serves as a highly significant predictor of their later BCT PR performance. In addition, parental education levels, household income, enrollment in private schools, and urban residency all exert significant positive effects on student outcomes, while household economic disruptions are associated with negative impacts. Educational investments during the ninth grade, including self-study time, school-based tutoring, and supplementary education, are also shown to significantly improve BCT scores.

The observed performance differences between public and private schools, as well as between urban and rural schools, are substantially reduced after controlling for background characteristics and learning investments. These findings suggest that such school-level differences can only partially explain academic outcomes. The deeper determinants of student performance lie in family background, early academic ability prior to junior high school, and educational efforts made during the ninth grade.