Meta-Analysis to Support an Integrated Theory of Multimedia Learning

该项目由教育和人力资源核心研究（ECR）项目支持，该项目支持STEM学习和学习环境的基础研究。越来越多的科学和数学信息通过多媒体呈现给学习者，例如讲解的教育动画或插图的超链接网站。教育媒体开发者使用的一些流行设计原则正在被新的研究所取代，需要新的原则来改进多媒体学习的设计、学习和未来的研究。这个大型综合项目将汇集500多项研究的趋势和发现，这些研究是对从中学到大学的学生进行的，他们通过多媒体学习科学或数学。一个主要的贡献是在使用多媒体学习时从“什么有用”转变为“什么对谁有用”。这个项目将会产生一本书、文章、研讨会和一个可搜索的网站。本项目将对多媒体学习进行meta分析，该meta分析是在一个新的多媒体学习元理论下组织的，该理论融合了几个现有的理论，包括个体差异（背景知识、工作记忆和空间技能）对刺激设计、任务设计和内容设计的影响的调节。使用文献检索的最佳实践，将对来自教育、心理学和其他学科的大约500项研究的影响进行编码，以确定对陈述性知识、推理、程序技能的影响，并将其转移到非指导内容。元回归方法将用于检查主要影响，与个体差异变量的相互作用，以及理论驱动和人口调节。研究结果应该为教育多媒体的设计者提供更细致的设计原则——尤其是自适应多媒体——并且应该通过解释这一新兴文献中相互矛盾的发现，为研究人员开辟重要的新途径。

The project is supported by the Education and Human Resources Core Research (ECR) program, which supports fundamental research in STEM learning and learning environments. Learners are increasingly presented with scientific and math information in multiple media, such as narrated educational animations or illustrated hyperlinked websites. Some popular design principles used by developers of educational media are being eclipsed by new research, and new principles are needed to improve design, learning, and future research on multimedia learning. This large synthesis project will pull together trends across findings from more than 500 studies conducted with students in middle school through undergraduate years, learning science or math from multimedia. A major contribution is in moving from "what works" to "what works for whom" when learning with multimedia. The project will result in a book, articles, workshops at conferences, and a searchable website.This project will involve conducting a meta-analysis of learning from multimedia that is organized under a new meta-theory of multimedia learning, which merges several existing theories and includes moderation of effects of stimulus design, task design, and content design by individual differences (background knowledge, working memory, and spatial skills). Using best practices for literature search, effects from approximately 500 studies across education, psychology, and other disciplines will be coded for effects on declarative knowledge, inference, procedural skills, and transfer to uninstructed content. A meta-regression approach will be used to examine main effects, interactions with individual difference variables, and both theory-driven and demographic moderators. Findings should result in more nuanced design principles for designers of educational multimedia- especially adaptive multimedia - and should open up important new avenues for researchers by explaining contradictory findings across this burgeoning literature.