| Two experiments were conducted to examine the use of direct-manipulation animation (DMA) to foster students' understanding of systems, a topic prominent in the National Science Education Standards. With DMA, learners manipulate the value of a parameter in a system while simultaneously observing how a change in one parameter triggers changes in other parameters. DMA incorporates the haptic channel in addition to the auditory and visual channels used in previous research. A model of cognitive processing in multimedia learning was proposed; accompanied by three hypotheses on the effect of using different instructional supports to learn about systems of varying complexities.;Both experiments investigated the effect of different instructional supports (i.e., medium, segmentation of instructional materials, and combination of external representation) on understanding mechanical energy transfer. Multiple measures were used to determine learners' comprehension and ability to extrapolate newly-learned information to reason about functional relationships between entities. The learning measures included: retention, "what-if scenarios" reasoning, problem-solving, and transfer tasks.;Experiment 1 examined the role of segmentation and type of instructional support for system understanding. Segmenting the process of energy transfer into multiple critical phases promoted better recall than depicting the system in a consolidated overview. A significant interaction was found in the transfer task: the animation group performed better with content presented in multiple phases; the picture group transferred newly-learned concepts to solve problems when content was displayed in a consolidated view.;Experiment 2 demonstrated that not all learners needed the same instructional support (text-only, text-and-pictures, and text-and-DMA) to understand simple, medium or complex systems. Seventh graders needed DMA with a simple text to reason about systems only with increasingly complex content; sixth graders required DMA with a simple text, regardless of system complexity.;Taken together, the findings suggest that increasingly complex systems-to-be-learned call for correspondingly sophisticated instructional supports. Further, the effectiveness of different supports may be a function of the intellectual demands of the domain-to-be-learned and learners' cognitive capabilities. Finally, the superior performance of the text-and-DMA group confirmed the efficacy of DMA as a congruent medium to provide crucial support that helps learners better understand dynamic functional relations between systemic entities. |
