一、內容：

　　本研究試圖找出有效能的英語教學實施，並檢驗能力分組跑班（簡稱跑班）介入英語低成就學生的可行性及成效。 第一部分：以觀課、訪問來描述八位由中央英語輔導團推薦優秀英語老師的教學。第二部分：新北市X、Y兩所國中各有兩班八年級進行「兩班三組」跑班，我們蒐集跑班、未跑班班級15名老師及713名學生的意見與英語文成就進展。第三部分：觀察宜蘭縣九所國中（886名學生）的AA-C跑班模式。 結果如下：

（一） 八名優秀老師的共同教學行為：1.班級經營及一般教學：建立清楚的增強系統及支持性班級氣氛、教學目標清楚、教學時間最大化、尊重個別差異。2.英語教學：老師以雙語（中、英）進行教學、明示教導語文技能（聽覺理解、文章唸讀、句型練習）、教導英文解碼（字母拼讀法）、看重進展監控、強調重複練習、提供豐富的英語環境與課外閱讀資源。

（二） 共有五次標準化成就測驗分數。A、B、C組學生的平均測驗分數達顯著差異，在X國中，A高於B，B高於C。「跑班」和「有效教學」對C組學生，則未有提升成就的效果，這可能與介入時機太晚有關。

（三） C組學生對分組跑班及英語學習的評價（兩次問卷的總分及平均值），顯著優於A、B組。Ｃ組學生即使認為跑班很丟臉，仍然希望下學期續辦。

（四） B組學生的學習狀況及對跑班的評價可能是三組中最低的。

二、教育政策之建議與貢獻：

（一） 兩班三組是可行的：老師們認為跑班利大於弊，它降低了班上學生能力的異質性，更能照顧每位學生，備課較容易，也認為跑班可以增進學生的學習動機和成就。

（二） 建議能力分組跑班從七年級開始，且要有彈性的進度與評量：部分八年級英語低成就生，英語程度與同儕差距已經過大，熟悉學校事務，而容易有行為問題。若能從七年級開始跑班，學生英語落差不致於太大，行為問題可能較少，將有助於教學的進行。老師們則希望有彈性的進度與評量，讓不同能力的學生能適性學習。

（三） AA-C跑班的可行性優於A-B-C跑班：AA-C跑班模式，只分兩個程度。可能受標籤化的學生人數降低，原班的文化及結構不致於有太大變動，不必擔心兩班學生混班後可能的不適應，可行性最高。A-B-C模式造成跑班學生增多，學生適應新班的問題增加，英文老師兼導師者的班級經營變得很困難。

（四） 要有強力的學校行政支持：多數師、生對跑班的抱怨，可以藉由行政支援減少的。例如，若學校可以提供C組較佳的教室、支持老師進行彈性的進度與評量、安排資深優良老師任教最弱的學生，都可能增加實施跑班的成功率。

1. Content：

　　The study attempts to identify delivery of effective instruction and to examine the effectiveness and feasibility of the ability-grouping intervention for low-achieving English learners in Taiwan. In the first part, eight outstanding teachers were recommended by the English-teaching counseling group in the government. These teachers were followed by classroom observation and interview to describe their teaching behavioral qualities. In the second, two junior high schools in New Taipei City administered the ability-grouping intervention that transformed two classes into three ability-based groups. Data-collection included the perception and progress in English achievement from 15 teachers and 713 students in the two schools. In the third, participating 9 schools in Yilan county adopted the AA-C ability-grouping (886 students). The results indicated:

(1) Common teaching qualities were recognized from the eight outstanding teachers. The first dimension was the qualities in classroom management and general instruction delivery, including the establishment of clear reinforcement mechanism, supporting class atmosphere, clear instructional objectives, maximum of instruction time, respect for indiviual difference. The second dimension was English-teaching qualities. The qualities were largely observable from bilingual instruction (i.e., Chinese-English), explicit instruction on language skills (i.e., listening comprehension, reading aloud, and sentence-pattern practice), decoding skills (i.e., phonics), monitoring of learning progress, emphasis on repeated practices, and easy access to stimulating environment and extensive reading materials for English learning.

(2) Five standardized tests were analyzed. All the tests showed a significant difference in the mean scores across Group A, Group B, and Group C. In School X, Group A showed the highest score, Group B the middle, and Group C the lowest. Regarding Group C, no significant difference was found between students who received ability-grouping intervention and those who did not. The non-significance may be related to the relatively late timing of the intervention.

(3) Group C assigned the highest appraisals of their English learning (i.e., the sum or mean scores in the two surveys), as compared to Group A and Group B. Even perceiving a sense of shame for ability-grouping, Group C hoped to continue ability-grouping in the next semester.

(4) Group B might express the lowest appraisals of both learning engagement and ability-grouping among the three groups.

2. Potential contributions and policy implications:

(1) The feasibility of transforming two regular classes into three ability-based groups was supported. The teachers perceived more strengths than weaknesses in such ability-grouping. They could address learning needs of each student more directly, and prepare the instruction with greater ease. They also perceived that ability-grouping could enhance students’ English learning motivation and achievement.

(2) Ability-grouping should proceed at Grade 7, and be accompanied by flexible course contents and assessment. Some of the low-achievers at Grade 8 had lagged far behind their peers in English proficiency, and had familiarized with school-related affairs, prompting more problem behaviors. If the ability-grouping proceeds at Grade 7, there may be a narrower gap in students’ English proficiency, as well as much fewer problem behaviors, which may ease the instruction delivery. The teachers hoped to have more flexibility in choosing course contents and assessment, enabling the adaptive learning of students at varying ability levels.

(3) The AA-C grouping has shown more feasibility than the A-B-C. The AA-C grouping, with two ability levels, thus decreased the number of students who might be labeled, brought a smaller change in the culture and structure of the original classes, and alleviated the concerns over regrouping maladaptation. These perceived strengths may contribute to a higher feasibility of the AA-C grouping. In contrast, the A-B-C grouping increased the number of participating students in regrouping, generated more problems in students’ adjustment to the new class, and placed more obstacles to the classroom management of teachers who served as both English and homeroom teachers.

(4) School administration should launch effective supporting measures. In fact, most students’ and teachers’ complaints for ability-grouping can be greatly minimized via supporting measures. For example, the possibility for success in the ability-grouping instruction can be largely raised, if school administration can provide a better classroom, enable teachers to deliver flexible course contents and assessment by their students’ English proficiency levels, and arrange the experienced outstanding teachers for the lowest-achievers in Group C.