簡易檢索 / 詳目顯示

研究生: 王為堯
Wang, Wei-Yao
論文名稱: 國中學生學習扶助篩選測驗成績對教育會考等第之預測研究:以金門地區某國中為例
Predicting Comprehensive Assessment Program Results Using Project for Implementation of Remedial Instruction Screening Test Scores:Taking A Junior High School in Kinmen as An Example
指導教授: 邱富源
Chiu, Fu-Yuan
口試委員: 江旭新
CHIANG, HSU-HSIN
黃敦煌
Duen-Huang Huang
學位類別: 碩士
Master
系所名稱: 竹師教育學院 - 教育與學習科技學系
Education and Learning Technology
論文出版年: 2025
畢業學年度: 113
語文別: 中文
論文頁數: 154
中文關鍵詞: 學習扶助篩選測驗國中教育會考區別分析預測研究
外文關鍵詞: Remedial Instruction Screening Test, Comprehensive Assessment Program, Discriminant Analysis, Predictive Research
相關次數: 點閱:10下載:0
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 本研究旨在探討國中學生七、八年級學習扶助篩選測驗成績對九年級教育會考等第的預測效果,以金門地區某國中108至113學年度的學生為研究對象。透過區別分析法,分析七年級、八年級及兩者組合的預測模型,評估其對國中教育會考國文、數學及英語等第的預測準確度,並計算最佳臨界切點,提供教育實務工作者及早識別學習困難學生的參考依據。
    研究結果顯示:
    (一)學習扶助篩選測驗對會考等第具有顯著預測力,三科組合模型的整體預測準確度依序為數學科(82.4%)、英語科(77.3%)及國文科(77.3%)。
    (二)各科模型對基礎等第(B)學生的預測準確度均高於90%,對待加強等第(C)學生的預測準確度則介於59%至82%之間,但對精熟等第(A)的預測皆為0%。
    (三)在時間效應上,除國文科外,八年級的篩選測驗普遍較七年級具有更佳的預測力。
    (四)組合模型相較單一年級模型具有更佳的預測效果,尤以數學科最為明顯。
    (五)各科篩選測驗的B-C等第臨界切點分別為:國文(七年級72分、八年級69分)、數學(七年級63分、八年級61分)及英語(七年級80分、八年級76分)。
    研究建議教育行政機關可將學習扶助篩選測驗納入會考成績預警機制,學校教師應善用篩選測驗結果及臨界切點進行早期識別與介入,並搭配多元評量,建立更完整的學生學習檔案,以提供適性化的教學支持。


    This study aims to investigate the effectiveness of the Project for Implementation of Remedial Instruction Screening Test scores in 7th and 8th grades for predicting Comprehensive Assessment Program (CAP) results in 9th grade. The research focuses on students from a junior high school in Kinmen from the 2019-2024 academic years. Using discriminant analysis, this study analyzes three predictive models (7th grade, 8th grade, and their combination) to evaluate their accuracy in predicting CAP levels in Chinese, Mathematics, and English, and calculates optimal critical cut-off points to provide educators with references for early identification of students with learning difficulties.
    The results show: (1) The screening tests have significant predictive power for CAP results, with overall prediction accuracy rates of 82.4% for Mathematics, 77.3% for English, and 77.3% for Chinese; (2) All models demonstrate high prediction accuracy for Basic-level (B) students (>90%), moderate accuracy for Below-Basic-level (C) students (59-82%), but fail to predict Proficient-level (A) students (0%); (3) Except for Chinese, 8th grade screening tests generally have better predictive power than 7th grade tests; (4) Combined models show improved predictive effects compared to single-grade models, particularly in Mathematics; (5) The B-C level critical cut-off points are: Chinese (72 points for 7th grade, 69 points for 8th grade), Mathematics (63 points for 7th grade, 61 points for 8th grade), and English (80 points for 7th grade, 76 points for 8th grade).
    The study recommends that educational authorities incorporate screening test results into CAP early warning mechanisms, and that teachers utilize these results and critical cut-off points for early identification and intervention, complemented by diverse assessments to establish more comprehensive student learning profiles.

    QR CODE