Design Factors for Effective Science Simulations: Representation of Information

Design Factors for Effective Science Simulations: Representation of Information

Jan L. Plass (New York University, USA), Bruce D. Homer (CUNY, USA), Catherine Milne (New York University, USA), Trace Jordan (New York University, USA), Slava Kalyuga (University of New South Wales, Australia), Minchi Kim (Purdue University, USA) and Hyunjeong Lee (University of Seoul, Korea)
DOI: 10.4018/978-1-60960-565-0.ch002
OnDemand PDF Download:


We argue that the effectiveness of simulations for science education depends on design features such as the type of representation chosen to depict key concepts. We hypothesize that the addition of iconic representations to simulations can help novice learners interpret the visual simulation interface and improve cognitive learning outcomes as well as learners’ self-efficacy. This hypothesis was tested in two experiments with high school chemistry students. The studies examined the effects of representation type (symbolic versus iconic), prior knowledge, and spatial ability on comprehension, knowledge transfer, and self-efficacy under low cognitive load (Study 1, N=80) and high cognitive load conditions (Study 2, N=91). Results supported our hypotheses that design features such as the addition of iconic representations can help scaffold students’ comprehension of science simulations, and that this effect was strongest for learners with low prior knowledge. Adding icons also improved learners’ general self-efficacy.
Chapter Preview

Learning From Animations And Simulations

There has been significant interest in the use of simulations and animations in education. Initial research was concerned with the comparison of the educational effectiveness of animations, i.e., visualizations that change over time, to that of static visualizations (Höffler & Leutner, 2007). Reviews of this research have not been able to identify overall benefits of animations over static pictures. Instead, they suggest that a more appropriate approach is to ask under what conditions and for whom one type of visualizations might be more effective than the other (Betrancourt, 2005). Other reviews have found that learner variables such as prior knowledge moderate the effectiveness of these representations, with low prior knowledge learners benefiting more from static images, and high prior knowledge learners benefiting more from dynamic visualizations (Kalyuga, 2006). In some cases, research has even shown that dynamic visualizations can interfere with learners’ performance of relevant cognitive processes, resulting in worse learning outcomes compared to non-animated, static visualizations (Schnotz, Böckler, & Grzondziel, 1999). There are also indications that the effectiveness of a particular animation depends on the design of the visualization, especially the appropriateness of the design for the specific learning goal and related tasks (Schnotz & Bannert, 2003).

In comparison to animations, which do not allow for significant user interactions, simulations can represent complex dynamic systems in which learners can manipulate parameters to explore and observe the behavior of a system (Gogg & Mott, 1993; Towne, 1995). This exploratory nature of simulations allows learners to engage in processes of scientific reasoning, i.e., problem definition, hypothesis generation, experimentation, observation and data interpretation (De Jong & van Joolingen, 1998; Kim & Hannafin, 2010a; Towne, 1995). Therefore, simulations have the potential to allow learners to understand scientific phenomena and transfer knowledge to novel situation better than other visual representations. In addition, while learners experience difficulties interpreting information from multiple representations such as text, pictures, and animation, the dynamic visualizations of system behavior in simulations can assist learners in interpreting concurrent changes in variables by revealing the underlying computational model (de Jong, 1991; Van der Meij & de Jong, 2006).

Although there are advantages to using simulations, learning from interactive simulations can also impose high cognitive load because some learners may not possess the required knowledge, cognitive abilities or metacognitive skills necessary to pursue scientific reasoning through simulations (Kim & Hannafin, 2010b; De Jong, & van Joolingen, 1998). Even though the integration of information in a simulation can help learners understanding the dynamic relationship between variables and representations, cognitive resources are required to relate the multiple changes that occur simultaneously in the various representations within a simulation (Van der Meij & de Jong, 2006). The integration of multiple representations may therefore result in high cognitive load (Lowe, 1999; Van der Meij & de Jong, 2006). Lowe (2003) attributes this added cognitive load to the changes in the form of representation, position of visual entities, and inclusion (appearance and disappearance) of visual components, which add to the visual complexity of simulations. Such a high degree of visual complexity may interfere with extraction and integration of relevant information from dynamic representation and incorporation of the information into learner’s prior knowledge (Lowe, 2003).

Complete Chapter List

Search this Book: