國小中、高年級閱讀理解歷程評量工具開發與研究

The development of reading comprehension assessment and research on students from fourth graders to sixth graders.

陳勁之(Chin-Chih Chen)

指導教授 ： 連啟舜

國立中正大學/教育學院/教育學研究所/碩士(2016年)

摘要

本研究旨在發展一套具有閱讀理論基礎的中文閱讀理解評量工具，藉此探討嘉義地區國小四、五、六年級學童以及各年級高、低能力學生在文本記憶、文本推理、知識整合與知識提取四個閱讀次歷程上的表現差異。本研究嘗試以Hannon & Daneman (2001) 閱讀成份作業之試題架構作為編製評量的基礎，並依據國小中、高年級學生之程度進行試題內容與形式的編寫及調整。在對181位四年級學生、184位五年級學生以及211位六年級學生進行正式施測後，本研究根據所蒐集之資料進行內部一致性信度分析、效標關聯效度分析、構念效度分析、重複量數分析、單因子變異數分析以及驗證性因素分析。研究結果顯示，本研究所編製之評量工具除了擁有良好的內部一致性信度，也有不錯的效標關聯效度與構念效度，其總分與「中文閱讀理解測驗」、「閱讀理解成長測驗」、「瑞文氏標準矩陣測驗 (SPM-P) 」以及「魏氏兒童智力測驗理解分測驗」四個效標測驗都達到中度相關，與各閱讀次歷程也有中、高度的相關。且透過重複量數的分析，三個年級的學生在新編閱讀理解測驗之次歷程總平均上也有差別，六年級的表現比四、五年級好，而四個閱讀次歷程在三個年級的學生成績上也有顯著區別。更進一步，依照單因子變異數分析的結果，各年級高、低能力學生在四個次歷程都有顯著差異，高能力學生的表現都優於低能力學生。最後，藉由驗證性因素分析，本研究之閱讀理解模型也符合RMR與RMSEA兩項絕對適配指標。

關鍵字

知識提取 ； 知識整合 ； 文本推理 ； 文本記憶 ； 閱讀理解

並列摘要

In this study, we tried to develop a theoretical-based Chinese reading comprehension assessment in order to explore fourth graders, fifth graders and sixth graders’ as well as high- and low-ability students from the three grades’ different performances on four sub-processes of reading comprehension, including text memory, text inference, knowledge integration and knowledge access. On the foundation of Hannon & Daneman (2001) component process task, we adjusted the content and the form of the task according to the students’ level of comprehension. After having the formal tests with 181 fourth graders, 184 fifth graders and 211 sixth graders, we put the data into statistical analysis, including internal consistency reliability, criterion-related validity, construct validity, repeated measures, one way ANOVA and Confirmatory Factor Analysis. The result showed that our new assessment not only had good internal consistency reliability but also fine criterion-related validity and construct validity. The whole score of the new assessment had medium correlations with our four criterion-referenced tests, including Chinese Reading Comprehension Test (Lin & Chi, 1999), Reading Growth Test (Su, 2015), Raven's Standard Progressive Matrices (SPM-P) and WISC-comprehension test , medium to high correlation with each reading sub-process as well. Through repeated measure, we found that the total average of our new reading comprehension assessment in each grade had discrepancy, sixth graders performed better than fourth graders and fifth graders. As for the four sub-processes in the three grades, there was also statistical significance. Furthermore, according to the result of one-way ANOVA, high- and low-ability students in each grade had statistical significance in the four sub-processes, high-ability students performed better than low ability-students. Finally, from the result of CFA, our reading comprehension models also met two absolute fit measures, RMR and RMESA.

並列關鍵字

knowledge integration ； text memory ； text inference ； reading comprehension ； knowledge access

參考文獻

林世華、黃寶園（1998）。柯氏性格量表效度概化之統合分析。師大學報：教育類，43 (2)，21-42。

連啟舜、陳弘輝、曾玉村（2016）。閱讀摘要歷程探究及閱讀理解歷程評量。

林怡君、張麗麗、陸怡琮（2013）。Rasch 模式建置國小高年級閱讀理解測驗。教育心理學報。45（1），39-61。doi: 10.6251/BEP.20121128

沈欣怡、蘇宜芬（2011）。推論性問題引導課程對國小四年級學童推論理解與閱讀理解能力之影響。教育心理學報，43，p.337~356

連啟舜（2002）。國內閱讀理解教學研究成效之統合分析研究。國立臺灣師範大學碩士論文。未出版，臺北市。