虛擬教具教學對學生學習成效之後設分析

Abstract

本研究之目的乃透過後設分析 (meta-analysis)，對國內的「虛擬教具教學」過去所累積的實證研究資料做分析，以瞭解它對於學生數學學習成效的影響，並進一步探討影響學習成效之中介變項。 本研究共蒐集24 篇虛擬教具教學的實證研究，使用Hedges 與 Olkin (1985) 的技術來計算加權平均效果量ｄ+值，並以「類別模式」探討虛擬教具教學對學生學習成效影響之中介變項。研究結果立即學習成效的ｄ+值為0.326，學習保留成效的ｄ+值為0.513，顯示實施虛擬教具教學在立即學習成效及學習保留成效上較實施傳統教具教學得到較高的學習效果。出版形式、實驗地區、年級、抽樣方式、學科主題、教學實驗時間、評量工具信度、實驗組教學者、操作模式、教學實驗軟體、控制組比較項目、教學形式、預試基點及統計方式等14個變項為虛擬教具教學對學生數學立即學習成效影響的中介變項。最後依據研究結果，研究者提出若干建議，供未來有志從事虛擬教具教學及後設分析研究之研究者參考，冀望能讓虛擬教具在教學上發揮最大的功效。

The purpose of this study is to explore the effects of virtual manipulative instruction on students’ academic achievement. Twenty-four studies were examined. The quantitative data of the collected studies were transformed into effect size (ES). The effect size was calculated by Hedges and Olkin’s (1985) procedure. Eighteen studied variables were analyzed by ‘Categorical Model ‘ to find moderator variables. The meta-analysis led to the following conclusions: 1. The whole effects suggest that virtual manipulative instruction is much more effective than the control group on students’ academic achievement. The effect size of learning effect is 0.326 and the effect size of retentive learning effect is 0.513. 2. There were 14 moderator variables of learning effects.They are publication source, teaching regions, students’ age, types of experimental control, topics of subject, duration of instruction, the ability of assessment, experimental teacher, operation mode, experimental software, the teaching disposal of control group, teaching time,base of test, quantification of outcomes. Based on the results of the study, suggestions for teaching and future research were discussed at the end.