Effectiveness of primary school students' reading comprehension on diverse expository prose formations

Abstract

隨着閱讀研究的蓬勃興盛，許多學者從認知層面探討人類閱讀的歷程，本研究主要目的欲運用認知負荷的設計原則(分心效應與冗餘效應)，編排說明文體，探討學生閱讀後的篇章理解表現與編排後的說明文體是否能降低學生的認知負荷。

本研究主要採用實驗研究，進行三次實驗。受試者為香港某私立小學四年級學童，受試者依據其語文理解測驗(前測)分數，隨機分派到實驗組與控制組。研究工具主要以閱讀理解表現測驗和認知負荷量表為主，訪談為輔。每次實驗大約35 分鐘，之後，研究者隨機抽取3-5 位受試者進行訪談。本研究主要以組別為自變項，閱讀理解表現和認知負荷量表的得分為依變項，前測分數為共變項，對收集到的資料進行單因子共變數分析。

資料分析顯示：實驗一(字、字義整合)與實驗三(文字、圖片整合)兩項實驗結果呈現，兩個組別並沒有因為不同編排而在閱讀理解與減輕認知負荷上有所差異；而在實驗二(段、段意整合)，兩個組別因不同編排的文本而在閱讀理解表現與認知負荷量表上呈現差異。最後，研究者亦提供教學與研究上的建議給教育學者和未來研究者。

With the prevailing research on reading around the world, many scholars have being explored the process of human’s reading in aspect of cognition. The main goal of this study was to investigate the effectiveness of primary school students’ reading comprehension on diverse expository prose formations which was designed based on the principles of cognitive load theory (the split-attention effect and the redundancy effect).

This study had three experiments and the fourth grade students were selected as subjects who randomly assigned to experimental group and control group based on their pre-test scores. The research tools included reading comprehension test, cognitive load scale (a 7-point Likert scale), and interview. Each test lasted 35 minutes, and then, 3-5 subjects were chosen to be interviewees. ANCOVA was used to analyze the collected data: group was the independent variable, the score on reading comprehension test and cognitive load scale were dependent variables, and the score on pre-test was covariate variables.

The findings showed that there were no significant differences between two groups in the first experiment (text, vocabulary definitions integration) and the third experiment (text, pictures/photos integration). However, there was a significant difference between two groups in the second experiment (text, text summary integration). Finally, the recommendations of teaching and research for educators and future researchers were provided as well.