國中階段數學科補救教學方案學習成效個案研究

隨著環境快速變遷，臺灣國中學生學業成就表現兩極化的問題日趨嚴重，教育機關嘗試以「補救教學」方案進行解決，並對補救教學成效多所讚揚。然而，「補救教學」究竟如何提升學生的學習成效、減少學習成就兩極化的問題，一直是身為國中教師的作者想認識的一項議題。為了探討國中階段數學科補救教學對學生學習成效的影響，本文針對一間位於大台北地區的國民中學，以三個入學年度之學生為對象，整理並分析他們三年期間的學習資料。應用「傾向分數配對法」，本文先找出與方案參與學生在特質及前測成績相似的配對學生。控制可能影響因素後，計算參與學生和配對學生後測成績差異，作為衡量補救教學對學生學習成效影響程度的基準。此外，本文亦衡量同一學期內的「短期效果」、相隔一年後的「長期效果」、「初次參與的效果」和「多次參與的效果」四個面向。透過資料的整理，本研究發現符合補救教學資格學生以男生為主、家庭經濟狀況較弱勢，且新移民子女及原住民族學生的比率為高。整理統計結果，本文呈現出兩項重要發現。第一，不同面向都出現負向結果的期數皆多於正向結果的期數。第二，具有統計的顯著差異的期數全為負向結果且皆發生在下學期，其中只有「2013八上初次參與組」發生在上學期。再以「學期」觀察長期與短期效果，本文也得到相類似的結果。首先，七年級上學期沒有負向結果的期別出現，而到了八年級下學期時則沒有正向結果的期別出現。此外，上學期正向結果的期數多於負向的結果，但下學期負向結果的期數卻遠高於正向結果。最後，具有統計的顯著差異皆發生在所有期別的「下學期」，而所有期別的「上學期」並沒有出現任何統計上的顯著差異。上述的統計結果和過去文獻呈現多處衝突，因此研究者訪談三位星星國中的數學教師，探究「顯著負向結果的理由」、「下學期較頻繁出現負向結果的原因」、以及「隨年級增長正負向關係逆轉」的議題。訪談結果不僅證明了統計結果的有效性，更發現顯著負向結果可能與「課程概念」和「考試難度」有關，間接說明補救教學的學習成效落差與學生的「學習態度與意願」、教師的「負向觀感」和家長的「支持程度」有所關聯。此外，下學期補救教學的多次失敗除了吻合過去文獻的「馬太效應」外，也重複說明學生的「學習態度」對學習成效的影響，這邊也證明「早期補救」的確對學生學習成效有所影響，亦即越早進入補救教學的學生，確實更有可能改善或維持其成績的表現。對於上述的結果，本文認為：第一，「早期介入」、「強制機制的設計」和「學習成效衡量機制與妥善使用」將是未來實務工作者能夠持續努力和關注的焦點。針對參與人數過少所造成的偏誤，或許能採用「合成控制法」來處理。第二，若研究者有足夠的經費與時間，應嘗試與特定學校進行「縱貫」的長時操作，更能控制各個可能影響的因子。此外，受限於資料的完整性，本文並未控制「學習動機」或「學習興趣」等因素，未來研究者或許納入這兩個變項進行處理。最後，本文只利用單一學校進行分析，結果僅適用單一學校，無法推論至整體方案的成敗，或許研究者在描述方案時能以更細緻的方式進行處理。對於未來有意進行類似研究的研究者，或許需要收集更多對應資料，用以認識研究學校的方案，找出更適當的衡量工具，以期真正衡量補救教學對學生學習成效上的影響狀況，提升參與學生改變的可能。

關鍵字

國中補救教學；國民中學；傾向分數配對法；補救教學；數學科補救教學；學習成效

並列摘要

As the society changes, polarization of academic achievement among Taiwanese junior high school students has become significant. Ministry of Education in Taiwan has tried to solve the problem with "Remedial Teaching", and praised its success. However, as a teacher in a junior high school, I am motivated to research on how "Remedial Teaching" improves students' learning outcomes and reduce the polarization of academic achievement. In order to explore students’ learning effectiveness of Remedial Teaching Program in a junior high school, this paper aims to organize and analyze a long-term study, including three entrance years of students and their comprehensive data in a selected junior high school located in the Greater Taipei area. Using "Propensity Score Matching" as a research tool, the author found out the "matching group" with similar scores and traits in the pre-test who is similar to participants. After controlling for possible confounding factors, the study compares the differences of test grades between program participators and the matching group to assess the impact of remedial teaching on student learning outcomes. And the impact of the remedial teaching are evaluated in four distinct dimensions, including "short-term effect" of the same semester, "long-term effect" of the outcome after one year , "first participation effect" and "multiple participation effect". The program evaluation reveals some profound findings. Above all, this study found that students who qualify for remedial teaching in Star junior high school(SJHS) are mainly male, the family economic situation is relatively poor, and the ratio of immigrant and aboriginal students is high. Regarding the impact of the remedial teaching program, there are two major statistical results. First, regardless of dimensions, the number of periods of negative results is more than that of positive results. Second, the number of periods with statistically significant differences is all negative and all occurs in the second semester, except for the “2013 First Participation Group” which occurred in the first semester. Regarding the results of long-term and short-term effects in the view of "semester", I get similar results. First, the first semester of Seventh Grade has no negative results, but there is no positive outcome in the second semester of Eighth Grade. Besides, the number of period of outcomes is more positive in all first semesters, but more negative in all second semesters. Finally, all the statistically significant differences occurred in the "second semester", and there was no statistically significant difference in the "first semester". Owing to the conflicts between the statistical results and the past literature, the researchers interviewed three mathematics teachers of SHJH to explore the reasons for the findings. In the interview, in addition to proving the validity of the statistical results, it was found that the significant negative results may be related to the "course concept" and the "difficulty of the examination", indirectly indicating the difference in learning outcomes of remedial teaching may result from the "learning attitude and willingness" of the students, the teachers' "negative perception", and the "support level" of parents. In addition, the repeated failures of the second semester remedial teaching are in compliance with "Matthew effect" and the impact of the student's "learning attitude" on learning outcomes of the past literature. The results prove that the "Early Remedy" do pose a positive impact on students' learning, and the students are more likely to improve or maintain their learning outcomes if they participant remedial teaching earlier. Using the results above, this paper argues that "early intervention", "design of enforcement mechanism" and "learning effectiveness measurement mechanism and proper use" will be the focus of future practical workers' continuous efforts and attention. The bias caused by the small number of participants may be treated by the "Synthetic Control Method". Second, if the researcher has sufficient funds and time, he or she should try to carry out long-term operations with specific schools and control the factors that may affect each. Besides, due to the integrity of the data, this article does not control factors such as "learning motivation" or "learning interest". Future researchers may include these two variables. Finally, this paper focuses on only one school, so the result can only be applied to this school. It is impossible to infer the success or failure of the entire program. Perhaps the researchers can deal with the plan in a more detailed way. For researchers who intend to conduct similar research in the future, t they may still have to collect more relevant information for a full understanding of the research school's program, and also involve more appropriate measurement tools to study the impact of remedial teaching on students' learning outcomes, making it possible to improve students’ condition.