Discovery learning, representation, and explanation within a computer-based simulation: finding the right mix

https://doi.org/10.1016/j.learninstruc.2004.06.008Get rights and content

Abstract

The purpose of this research was to explore how adult users interact and learn during an interactive computer-based simulation supplemented with brief multimedia explanations of the content. A total of 52 college students interacted with a computer-based simulation of Newton’s laws of motion in which they had control over the motion of a simple screen object—an animated ball. Two simulation conditions were studied, each differing in how the feedback of the ball’s speed, direction, and position was represented: graphical feedback consisted of animated graphics and textual feedback consisted of numeric displays. In addition, half of the participants were given simulations supplemented with brief multimedia explanations of the content modeled by the simulation in order to investigate how to promote referential processing, a key component of dual coding theory. Results showed significant differences for both the use of the explanations and simulations containing graphical feedback in helping participants gain both implicit and explicit understanding of the science principles.

Introduction

Physics has long been regarded as one of the most abstract and difficult subjects to learn. A common belief is that the students who do well in physics have a special aptitude for learning science and mathematics. It is easy to understand how many might believe that the only students who are going to “need” physics are future engineers and physicists. However, instead of reserving physics for a gifted few, many feel that technology offers the chance to change commonly held perceptions about who can or should learn physics (White & Frederiksen, 1998). For example, complex systems such as physics can easily be modeled on desktop computers. Computer simulations make complex systems accessible for students of varying ages, abilities, and learning levels. The computer, instead of the student, can assume responsibility of processing the underlying mathematics in order to let the student begin exploring a complex system by first focusing on conceptual understanding (de Jong and van Joolingen, 1998, Penner, 2000/2001, Roschelle, Kaput and Stroup, 2000). The emphasis of this approach is on experiences, rather than explanations, of a domain. Of course, this is not to suggest that explanations are not important. Indeed, much current research points to the advantages of multimedia explanations for learning (Mayer, 2001). However, the interactive affordances of computer simulations and other modeling environments suggest an educational model based on experiential learning. The role, timing, and influence of explanations while learning science principles need to be reconsidered and reexamined.

Despite the enthusiasm for educational simulations, many challenges to the effective design of simulations remain. Among the most difficult are questions about the design of a simulation’s interface (Edwards and Holland, 1992, Schneiderman, 1998). For example, designers currently can include a wide range of visual, textual, and aural elements in the development of simulations. As the range of available design options increases, so too does the complexity of design decisions. One of the most important considerations in a simulation’s interface design is how to provide meaningful feedback to the user. Cognitive psychologists have long regarded feedback as a critical source of information to assist learners in restructuring their knowledge and supporting their metacognitive processes (Bransford and Brown, 1999, Kulhavy and Wager, 1993). Given the range of ways computers can represent feedback in a simulation, research is needed to ensure that design decisions are made based on the psychological needs of the individual user and not simply on what the computer is capable of doing.

Much research demonstrates that the way information is represented matters greatly in the learning process, at least for memory tasks (Clark and Paivio, 1991, Paivio, 1990, Paivio, 1991, Sadoski and Paivio, 2001). Research indicates that pictures are superior to words for remembering concrete concepts (Sadoski & Paivio, 2001). Among the various theories that have been proposed to explain this, Paivio’s dual coding theory appears to have the strongest empirical support (Anderson, 1978, Paivio, 1990, Paivio, 1991, Sadoski and Paivio, 2001). Dual coding theory divides cognition into two processing systems—one visual and one verbal. Although the research supporting dual coding theory is based almost exclusively on evidence derived from supplementing printed text with visuals (including paired associative tasks) (e.g., Sadoski, Goetz and Avila, 1995, Sadoski and Paivio, 1994), this theory also holds promise in guiding research in computer-based multimedia environments (Mayer, 2001, Mayer and Sims, 1994, Rieber, 1996).

Dual coding theory predicts three separate levels of processing within and between the visual and verbal systems: representational, associative, and referential. Representational structures (either visual or verbal) are formed depending on the nature of incoming information (i.e., visual and verbal information from the environment triggers the visual and verbal systems, respectively). Associative processing leads to connections constructed within either the visual or verbal systems, whereas referential processing leads to connections made between the visual and verbal systems. Referential processing is particularly important because dual coding theory predicts that learning will be enhanced when information is encoded in both systems (i.e., dually coded). Information that is dually coded has twice the chance to be retrieved and used (Kobayashi, 1986). Instruction that promotes dual coding has obvious advantages.

Our past research has shown that the way feedback is represented also matters when learning from simulations of physical science concepts and principles (i.e., laws of motion). Participants increased their implicit knowledge of physics when they interacted with a physics simulation given graphical feedback, but they were unable to demonstrate increased explicit understanding based on the way the feedback was represented (Rieber, 1996, Rieber and Noah, 1997, Rieber, Noah and Nolan, 1998, Rieber, Smith, Al-Ghafry, Strickland, Chu and Spahi, 1996). Implicit understanding was measured by participants’ performance in a game-like activity whereas explicit understanding was measured using a traditional performance test (i.e., multiple-choice question format). The increase in implicit learning given graphical feedback indicated that representational and associative processing occurred almost exclusively within the visual system. Participants’ difficulty in acquiring explicit understanding of the physics principles modeled by the computer was attributed to the highly interactive nature of the discovery-based simulation. Simulations that model physical phenomena (such as physical science) may not provide the learner with sufficient time or guidance for interpreting the continual stream of feedback by both the visual and verbal systems. In other words, the “video game-like” quality of the simulation may have interfered with referential processing by only promoting processing in the visual system and discouraging processing in the verbal system.

The purpose of this study was to investigate ways to facilitate or enhance referential processing as a user interacts with a computer simulation. Our previous research used a pure discovery-based approach—no instruction was included or embedded in the simulation. While highly experiential open-ended simulations appear beneficial in many ways, they do not seem to adequately promote reflection of the science principles. Reflection is an important component for referential processing. One way to promote reflection is to provide the student with a brief explanation of the scientific principle being modeled by the simulation after the student has had an opportunity to interact with the simulation, but not yet master the content of the simulation. In this study, five short multimedia explanations (brief text and simple animation) were embedded throughout the simulation. Each multimedia explanation focused on one of the fundamental physics principles modeled in the simulation. Such a use of alternative representations is consistent with the research on multiple representations (Ainsworth, 1999). Multimedia explanations, such as these, serve to complement the student’s simulation experience by focusing attention on the specific scientific principle at work. The multimedia explanations should also help to constrain the student’s interpretation of their experience with the interactive simulation so as to help them focus on only the most germane principles of the physics being modeled. Hence, these brief explanations should help the student to organize their interactive experiences by helping them to make meaning from the feedback generated by the simulation, an essential step on the road to understanding (Mayer, 1989).

It was hypothesized that supplementing the simulation with multimedia explanations of the content would facilitate all three types of processing predicted by dual coding theory for explicit learning. Also, since previous research suggests the apparent dominance of the visual system during a user’s interaction with simulations similar to these, it was also hypothesized that the embedded explanations would promote more referential processing when participants were given graphical instead of textual feedback.

Several data sources were used in this study. Traditional performance measures (e.g., question-based pretests and posttests) were used to assess participants’ explicit understanding of the science principles modeled in the simulation. However, such formal tests do not assess other levels of understanding that are embedded in a task. For example, bringing a car to a smooth controlled stop requires an extensive understanding of many motion principles. However, this understanding remains situated in the act of driving—the individual may not be able to explicitly describe the physical relationships at work. Although we recognize that a learner’s ability to transfer conceptual understanding from one task to another (such as to a posttest) remains an important indication of learning, this study also used a measure of implicit understanding found useful in our earlier research. Participants were asked to complete the simulation in a game-like context. Since an understanding of the motion principles is necessary to be successful at the game, the game score provides an alternative data source useful when compared with the participants’ scores on traditional performance measures.

Section snippets

Participants

A total of 52 junior and senior undergraduate students participated in this study. These participants were enrolled in an introductory computer education course. Participation was voluntary, though extra credit in the course was provided to students as incentive to participate. The average age of all the participants was 21.6 years. The participants were predominately female (90.3%).

Materials

The materials consisted of a computer-based simulation of Newton’s laws of motion. Participants had direct

Pretest/posttest of principle learning

Percentage means and standard deviations are contained in Table 1. A significant interaction was found between the pretest/posttest of principle learning and explanation, F(1, 48)=9.55, p<0.01, MSerror=190.51. The difference between the pretest and posttest scores was greatest when participants were provided with embedded explanations than when they were not given the explanations. There was also a significant interaction found between the pretest/posttest of principle learning and feedback

Discussion

The purpose of this research was to investigate ways to facilitate or enhance an individual’s learning of physics principles while interacting with a computer simulation based on an experiential approach. Many facets of learning were studied in this research, such as implicit versus explicit understanding as well as patterns of interactivity and frustration. While computers afford the design of highly interactive open-ended learning environments such as simulations, decisions about how to

References (36)

  • R. Clark

    Antagonism between achievement and enjoyment in ATI studies

    Educational Psychologist

    (1982)
  • T. de Jong et al.

    Scientific discovery learning with computer simulations of conceptual domains

    Review of Educational Research

    (1998)
  • J. Dempsey et al.

    Since Malone’s theory of intrinsically motivating instruction: what’s the score in the gaming literature?

    Journal of Educational Technology Systems

    (1993 1994)
  • W. Dick et al.

    The systematic design of instruction

    (2001)
  • S. Kobayashi

    Theoretical issues concerning superiority of pictures over words and sentences in memory

    Perceptual and Motor Skills

    (1986)
  • R.W. Kulhavy et al.

    Feedback in programmed instruction: historical context and implications for practice

  • M.R. Lepper et al.

    Intrinsic motivation and instructional effectiveness in computer-based education

  • Cited by (105)

    • A head mounted augmented reality design practice for maintenance assembly: Toward meeting perceptual and cognitive needs of AR users

      2022, Applied Ergonomics
      Citation Excerpt :

      Notwithstanding the fact that the dynamic visual modality of video or animation has been agreed to be engaging and efficient for information communication, the difference between these two in terms of how users interact with different types of dynamic information is still unclear. Some research suggest that animation might be more engaging and emphasizing the components of the process to learn (Rieber et al., 2004), other research suggest that visualisations might distract from the actual process being learnt (Lowe, 1999, 2004) suggesting that videos might be a more effective media for learning. Furthermore, visual modalities have the advantage of simplifying complex information, easing comprehension of the process and illustrating how different aspects of the process relate to each other (Smaldino, 2012).

    • Instructional support for learning with agent-based simulations: A tale of vicarious and guided exploration learning approaches

      2019, Computers and Education
      Citation Excerpt :

      In fact, regarding exploration of agent-based models, it is still unclear how to organize and design instructional guidance for different learner populations. On one hand, agent-based models are exploratory models designed so that students can discover things by experimenting within a microworld; on the other hand, without providing explanations, understanding of complex systems won't become explicit but might remain implicit (Rieber, Tzeng, & Tribble, 2004). Vitale, McBride, and Linn (2016) compared two forms of automated guidance, direct explanations and prompts to motivate learners (knowledge integration guidance), to support learning of complex systems with agent-based models.

    • Investigating the Effect of Simulator Functional Fidelity and Personalized Feedback on Central Venous Catheterization Training

      2018, Journal of Surgical Education
      Citation Excerpt :

      Finally, prior work by the authors has shown that participants improved their performance throughout the course of training on the DHRT system and they approached the level of an expert at the end of their training.19 An advantage of using functionally high-fidelity simulators like the DHRT for IJ CVC trainings is that they can be designed to automatically present feedback in both textual and graphical forms which can improve information retention.20 For example, the DHRT system provides users with feedback after each insertion trial on the needle angle, final position of the needle tip, number of insertions, and amount of unnecessary movements, see Yovanoff et al.21 for discussion.

    • Six instructional approaches supported in AIED systems

      2023, Handbook of Artificial Intelligence in Education
    View all citing articles on Scopus
    View full text