本研究探究可預測書讀寫整合教學及一般生字與句型教學，對臺北市小五生英語寫作能力之影響。研究問題為：一、就詞彙及句法能力來探究可預測書讀寫整合教學是否可提升小五生英語寫作能力？二、可預測書讀寫整合教學(實驗組)及一般生字與句型教學(控制組)，何者較能提升小五生英語寫作能力？研究者對北市六十八位小五學生施以劍橋兒童CYLET聽力及閱讀寫作前測同質性測驗，分實驗組與控制組各三十四人，分別施以可預測書讀寫整合教學及一般生字與句型教學。實驗教學進行十四週，每週兩次，每次二十分鐘。實驗組與控制組於實驗教學前後各接受一次自編「英語寫作能力測驗」。研究者透過CHILDES進行前後測資料編碼與分析，以詞彙的運用及句法能力為評斷指標。另以相依樣本t檢定分別檢驗兩組研究對象前後測的表現是否顯著提升，並以獨立樣本單因子共變數分析比較兩組前後測成績。結果顯示，可預測書讀寫整合活動，能有效提升小五生英語寫作能力，特別是詞彙的運用；句法能力上，對提升句子總數與正確句總數有顯著效果，但對提升平均句長的教學效果則有限。實驗組可預測書讀寫整合教學之成效優於控制組，控制組後測在詞彙的運用及句法能力表現皆退步，其中句子總數退步幅度達顯著差異。據此，研究者提出教學及研究建議。

This study aimed to understand and compare the effects of integrated reading-writing instruction with English predictable books and traditional vocabulary and sentence patterns instruction on the elementary school fifth graders’ English writing ability. Sixty-eight 5th graders participated in the study for 14 weeks, two 20-minute sections per week; 34 people were assigned to the experimental group and 34 the control group, with the homogeneity in their CYLET between two groups. The experimental group received integrated reading-writing instruction with predictable books, while the control group received traditional vocabulary and sentence patterns instruction. Pre- and post-tests on writing were conducted before and after the experimental teaching. CHILDES was used to analyze English vocabulary and syntax as shown in the participants’ writing performances. Paired-Sample t-Test and One-Way ANCOVA were used for statistical analysis. The results showed that integrated reading-writing instruction with English predictable books effectively enhanced the participants’ writing. The experimental group improved significantly in the posttest as compared to the control group, especially on their English vocabulary use in writing. Traditional vocabulary and sentence patterns instruction failed to improve the control group’s English vocabulary and syntax as shown in the writing performances. The researchers thus made recommendations for teaching and future research.