Self-Regulation of Learning with Multiple Representations in Hypermedia

  • Authors:

  • Jennifer Cromley;Roger Azevedo;Evan Olson

  • Affiliations:

  • Department of Human Development and Cognition and Technology Laboratory, University of Maryland, 3304 Benjamin Building, College Park, MD 20742, USA, jcromley@umd.edu, razevedo@umd.edu, eolson01@y ...;Department of Human Development and Cognition and Technology Laboratory, University of Maryland, 3304 Benjamin Building, College Park, MD 20742, USA, jcromley@umd.edu, razevedo@umd.edu, eolson01@y ...;Department of Human Development and Cognition and Technology Laboratory, University of Maryland, 3304 Benjamin Building, College Park, MD 20742, USA, jcromley@umd.edu, razevedo@umd.edu, eolson01@y ...

  • Venue:
  • Proceedings of the 2005 conference on Artificial Intelligence in Education: Supporting Learning through Intelligent and Socially Informed Technology
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

A body of research has demonstrated when multiple representations of content help students learn. Few studies, however, have used process measures to understand what different cognitive processes students enact when learning from different representations. We collected pretest, posttest, think-aloud, and video data from 21 undergraduate students learning about the human circulatory system using a hypermedia encyclopedia. We measured learning as a change in a participant's mental model of the circulatory system from pretest to posttest. Students who learned more tended to spend less time in Text. While viewing Text alone, amount of learning was most strongly associated with verbalizing a smaller proportion of Feeling of Knowing, Free Search, and Selecting a New Informational Source. For Text + Diagrams, the amount of learning was most strongly associated with verbalizing a larger proportion of Inference and Self-Questioning. For Animation, the only significant variable was Summarizing. When not using the hypermedia environment, the significant variables were Feeling of Knowing, Prior Knowledge Activation, and Taking Notes. We close with implications for designing hypermedia environments for learning about complex science topics.