不考試，公平嗎？以全國考招資料檢視多元入學公平性

Fairness of Nontraditional Paths to Higher Education: An Analysis of Administrative Data

教育機會公平在國內外教育社會學領域中，始終有著高度重要性。臺灣則在大學多元入學制度之爭議脈絡下，入學公平議題被反覆檢視。多數輿論認為第二階段篩選不利於資源貧弱學生。然而既有研究受限於資料型態，僅能觀察「已被錄取並就讀大學者」。本研究運用2014至2020年度之大考中心及招聯會串接資料（亦即108課綱實施之前的應屆考生）進行加值應用分析，並利用此類型資料之優勢，將「個人申請入學」的兩個篩選階段區分開來─第一階段（學測分數）與第二階段（書審、面試），以突破既有限制。以下為精簡研究結論：(1)對所有觀察年份來說，中低／低收入戶學生之一階通過率，均低於一般家戶學生。但到了二階，此差異變小，且常出現「經濟弱勢學生之通過率高於一般學生」。(2)控制年份與學校之後，經濟弱勢生一階通過的「勝算」（odds）比一般學生少了18%。這些學生之所以於一階篩選不利，較低的學測分數為主要解釋因素。(3)控制年份與學校後，經濟弱勢生之二階通過率並沒有比一般學生低。進一步考慮學測，在相同級分的條件下，經濟弱勢生於二階獲得正取之勝算要比一般收入戶考生高出了15-16%。本研究顯示：看似六親不認的考試，事實上是造成優勢與弱勢家庭學生之間入學機會嚴重落差的篩選工具；大眾擔憂不夠客觀的二階書審及面試，反而可能成為家庭資源匱乏的學生之翻身管道。

The 2002 Multiple Channels of College Entrance policy supported nontraditional paths to higher education for Taiwanese students, but the general public has heatedly debated is fairness. Three important findings emerged from an analysis of administrative data from the General Scholastic Ability Test (GSAT) and 2014-2020 admissions data (prior to the introduction of the Education Ministry’s “108 new curriculums”):
Low-income students had significantly smaller passing rates during stageone selection (GSAT), but the disadvantage was much less during stage-two selection, depending on application packages and college interviews.
After controlling for number of years and schools, the odds of low-income students passing the stage-one selection process were 18% lower than for other students, primarily due to their lower standardized test scores.
No major difference in passing rate was observed between lower-income and other applicants during the stage-two selection process. After controlling for GSAT scores, they actually had 15-16% greater chances of gaining college admission.