有效的中文閱讀理解策略：國內實徵研究之最佳證據整合 Strategy: Best-Evidence Synthesis of Taiwanese Empirical Studies

Abstract

以研究證據導引實務決策的理念愈來愈受國內、外學者重視，但證據來源愈趨多元時，其可信度也備受挑戰，因此，本研究採用實徵研究之最佳證據整合觀點，經累積及整合國內研究，以提出有效的中文閱讀理解策略。研究者利用數個關鍵字檢索國內主要文獻資料庫，就回傳資料逐篇比對納入與排除準則後，總計獲得57篇文獻，經整合，結果發現自我提問、教師提問－學生回應、故事結構分析、圖形組織繪製、摘要與交互教學等策略，皆對於一般中、小學學生具有一定程度閱讀理解教學成效。其中，摘要策略具有最高的教學效果，其平均效果量為＋0.58，而自我提問策略則相對具有最低教學效果，其平均效果量為＋0.35。整體而言，本研究所整合中文閱讀理解策略皆具有一定教學成效，值得未來實務推廣參考。 The adoption of evidence-based education has increased noticeably among Taiwanese practitioners and researchers. However, several accompanying concerns, such as the reliability and source of evidence, remain controversial. This study proposes an effective Chinese reading comprehension strategy from the perspective of best-evidence synthesis. After a few keywords were used to search Taiwanese bibliographic databases, 57 articles were coded and synthesized on the basis of inclusion and exclusion criteria. The results show that nearly all of the comprehension strategies were effective, including self-questioning, question and answering, story grammar, graphic organizer creation, summarizing, and reciprocal instruction. The highest average effect size was +0.58 for summary writing, and the lowest average effect size was +0.35 for self-questioning. Finally, some suggestions and implications are proposed for future studies.

Keywords

• 文本結構
• 交互教學
• 研究整合
• 提問
• 摘要
• text structure
• reciprocal instruction
• research synthesis
• questioning
• summarizing