虛擬教具教學對學生學習成效影響之後設分析

Abstract

本論文研製之目的在於透過後設分析法（meta-analysis），探討虛擬教具教學對學生數學學習成效之影響，並進一步分析影響學習成效的中介變項。本研究蒐集2000年至2016年間共27篇虛擬教具教學的實證研究，採用Hedges和Olkin(1985)以及Cohen(1988)的技術，透過CMA和Excel程式作為研究工具計算整體加權效果量，並以CMA的類別模式探討中介變項。本研究結果歸納如下： 一、立即學習成效的整體加權效果量為0.628，表示虛擬教具教學具有立即的數學學習成效，比傳統或實體教具教學有更好的效果。 二、學校位置、實驗對象之年級、學科主題、工具信度、實驗組教學者、教學形式等6個變項為虛擬教具教學對學生數學學習成效影響的中介變項，並得到以下六點發現，分述如下： 1.北部地區學生使用虛擬教具教學學習成效較好。 2.國小中年級實施虛擬教具教學對學生學習成效最好。 3.數學測量主題上實施虛擬教具教學對學生學習成效最好。 4.有進行信度考驗並提供數據的研究，會有較好的效果。 5.實驗組教學者為原授課教師非研究者，實驗組學生的學習成效較佳。 6.正常教學以外的時間使用虛擬教具教學，其學生學習成效較佳。 最後，根據本研究上述結果提出相關建議，俾供各級學校教師、虛擬教具研究者，以及後設分析研究者參考。

This study attempts to discuss, through meta-analysis, the influence of virtual manipulatives on students’ academic achievement (SAA), and further analyze the intervening variable that affects SAA. We collected 27 virtual manipulatives-related empirical researches from 2000 to 2016. The techniques of Hedges & Olkin (1985) and Cohen (1988) were employed to serve as research tools through CMA and Excel and calculate the overall weighted effect size. We discussed the intervening variable with a categorical model in CMA. Results are provided as follows: 1. The overall weighted effect size of the immediate learning outcome was 0.628. This suggests that virtual manipulative teaching has an immediate learning outcome in mathematic and outweighs traditional or physical manipulative teaching. 2. School location, grade of subjects, academic subject topics, ability of assessment, teacher in the test group, form of teaching – these 6 variables were the intervening variables through which virtual manipulative teaching affected SAA in math. Findings are listed below: 1) Students in the north who received virtual manipulative learning had a better SAA. 2) Practicing virtual manipulative teaching among middle-grade elementary school students had the best SAA. 3) Practicing virtual manipulative teaching in math measurement subjects had the best SAA. 4) An empirical study that conducted reliability test and provide data would have a better result. 5) When the teacher in the test group was the original instructor, students in the test group had a better SAA. 6) When virtual manipulative teaching was implemented outside of normal teaching sessions, its SAA was better. Finally, some suggestions are proposed based on the foregoing results for reference of school teachers at all levels, virtual manipulative researchers, as well as meta-analysis researchers.