Abstract
The EcoRaft Project is an interactive exhibit for teaching 8 to 12 year old children about restoration ecology. It uniquely approaches this educational task by using 3-D animations on desktop and tablet PCs to engage participants in a collaborative venture to restore graphically depicted ecosystems. Tablet PCs act as virtual rafts on which animated seeds and birds are carried between desktop PCs. The latter act as virtual islands in varying stages of deforestation.
Distilling information from complex content domains to be internalized by any target demographic is demanding. In the EcoRaft Project, the impact of complex data on the cognitive economies of participants is managed through the implementation of video, audio, and text techniques, as well as techniques of manipulating spatial orientations and social interactions. Research on how children multimodally resemiotize information was observed and employed to convey the most salient of the myriad details involved in both the ecological restoration process and the interactions of various organisms in a Costa Rican rainforest. The results of preliminary onsite investigations reveal that the project may be an effective pedagogical tool: EcoRaft extends beyond current cognitive theories on multimedia presentations, which predominately focus on 2-dimensional exhibits and agents, and demonstrates a novel paradigm for using 3-dimensional interactive exhibits in virtually augmented pedagogical settings.
Introduction
In “Simplicity & Complexity in Games of the Intellect,” biologist Lawrence B. Slobodkin asserts that one of the overarching principles behind adding or removing information from a simulation is to avoid the dullness of simplicity and the confusion arising from surfeit complexity:
Adding complexity to either the perceived world or a simplification of it must be done with a discrete purpose, lest complication become “obfuscation.” In creating a simplified or complicated system, the divergence from reality is acknowledged. To be simplistic or to obfuscate betrays misunderstanding, or worse, an attempt to hoodwink. [21]
The balance between obfuscation and over-simplicity is prominent in education and cognitive science literature, especially in conjunction with concerns over the limitations of working memory and cognitive load. In the face of these cognitive bounds, multimedia has been shown to be an effective didactic tool by way of disseminating information along multiple cognitive and sensory channels, with positive studies of multimedia in the natural sciences figuring prominently. [30] Although such studies tend to focus solely on visual and aural information, the use of mobile devices in augmented spaces, where virtual worlds commingle with real space, has been relatively unexplored as a didactic tool. In such spaces learners can engage in a social construction of knowledge while experiencing additional efficacious multimodal effects through haptic and proprioceptive stimuli.
Well grounded in the science of restoration ecology [2], the EcoRaft Project uses virtual environments to depict some of the flora and fauna of Costa Rican rainforests. Tablet PCs act as virtual rafts in a sea of real space and are maneuvered by participants to get plants and animals from thriving virtual islands to deforested ones. The islands are portrayed through animations on desktop PCs.
A novel orchestration of animation and networking techniques has made the EcoRaft Project an engaging model for enabling children to learn about elements of restoration ecology. The techniques include combining IrDA with TCP/IP networking to allow tablet PCs to share autonomous mobile code (as embodied agents) with desktop PCs, using sound to maintain the illusion that objects are transferred between islands and rafts, and providing novel interactions between embodied mobile agents and participants. It also employs cognitive, visual communication, and educational techniques and theories, applied along the guidelines of Richard E. Mayer’s Cognitive Theory of Multimedia Learning (CTML), to provide participants with an engaging learning experience. [13] But, EcoRaft extends CTML and its supporting theories into three dimensions, exploring ideas of multimodal learning and the resemiotization of information of its target demographic in a virtually-augmented three dimensional space.
Demographics
The targeted demographic for the project, children of ages 8 to 12, has been chosen in part because, like younger children, they tend to naturally resemiotize a kaleidoscope of available multimodal symbols in a transformatory fashion that not only favors their desire for self-expression, but also favors what they find attractive and relevant. [8, 14] This tighter focus for sense making suggests that the age group can comprehend the multimedia information the project presents with little distraction from what they consider to be irrelevant details. A task of a pedagogical method, then, is to discover and exploit what the targeted demographic is likely to key into.
Another factor to consider is that the 8 to 12 group is more likely than younger children to have some basic understanding of rainforests. The prior familiarity with the domain enables them to comprehend a carefully constructed multimedia presentation of the topic more easily. [29]
Agents
With the proliferation of cell phones, PDAs, laptops, and other computing devices, the adaptation of multimedia for learning through the interaction wjth mobile platforms is an interesting contemporary paradigm to explore. This is especially true when embodied mobile agents are used to make such learning more dynamic and, thus, more engaging. Increasing the level of engagement in a pedagogical arena facilitates learning. [3]
Related Work
The Cognitive Theory of Multimedia Learning
While this paper isn’t primarily about CTML, it is concerned with how efficaciously CTML techniques have been employed the EcoRaft Project.
Mayer’s theories have developed from work done by Sweller [22] and others on cognitive load theory and from theories supporting learner-centric instruction. The three principles that underlie CTML are that learning through two channels of perception are better than one, that working memory has a limited capacity, and that students actively process information as they construct knowledge. From these he has derived seven more principles he calls multimedia, spatial contiguity, temporal contiguity, coherence, modality, redundancy, and individual differences. The EcoRaft Project adheres to all seven, with the exception of the redundancy principle. According to this principle, using animation with narration and text is worse than using just animation and narration. The project violates this is in such a way as to alleviate any negative multimodal effects. How the project deals with the principles will be discussed further in the implementation section of this paper.
Arguments for constructivist theories of cognition are still prominent in the fields of education, visual communication, and cognitive science. Their lasting relevance has lead to many studies of how to apply them to multimedia learning.
According to Tudoreneau, the visualization of information extends cognitive resources by shifting the burden of internal representations of information to external media. [27] This supports the work by Sweller and others on cognitive load, which asserts that working memory has limited capacity, and the automation of information processing is more efficient than tasking the mind with developing schemas. This work also proffers that using dually modal sources of information improves learning by dividing the burden of developing schemas between two sensory channels. The use of schemata to get information from working memory into long-term memory is one of the main tenets of constructivist theories.
Lowe avers that adding motion (temporal change) may increase cognitive load by raising information processing demands and reducing the amount of relevant data that a participant consumes. Such temporal changes as transformations, translations, and transitions possibly increase the overall complexity of information while simultaneously directing the attention of subjects away from relevant material. This all results from a split-attention effect usually associated with combining text with animation. On the other hand, presentations in which participants have some control over the animations may have some salient effects on their construction of knowledge. [12]
On the other hand, animation alone is insufficient to effectively ensure the retention and transfer of information. Even with text added, animation may oversimplify the material and merely encourage mimicry by participants without fostering learning. [17] Additionally, researchers have had difficulty generalizing Mayer’s theory beyond his test cases. In some cases, animation with procedural text was used, in others descriptive text was used. In yet others, the quality of the animation could have been improved. [11, 23, 30]
CTML has yet to be refuted. The thrust of Mayer’s research has been with depictive diagrams, but most of the subsequent work challenging him has analyzed the use of descriptive diagrams. Also, CTML seems better suited to the natural sciences than the social sciences used in these other studies. [30] The ages of the human subjects of these researches might have been an issue too: children seem to learn better than adults with a combination of animation and text under certain conditions. [11]
The EcoRaft Project favors a constructivist approach to creating knowledge in the way it encourages children to learn about restoration ecology; it allows them to interact with the text, narration, and animations in a meaningful way, as well as with adults and other children. The very social aspect of exhibit invites cooperative learning between children of similar ages while allowing parents to provide scaffolding in the experience. The children are allowed to explore what the exhibit has to offer and develop their own knowledge while being gently guided to that knowledge by the structure the exhibit provides.
Agents
In the EcoRaft Project, mobile autonomous agents, code that can move between platforms and independently interact with their environment, are embodied in computer animated seeds and hummingbirds.
Mobile agents have been used for such tasks as network resource discovery [5] monitoring the status of the stock exchange [9], searching for the lowest prices of airplane tickets [16], and information dissemination and discovery for the military [7]. Mobile agents have been used in augmented reality scenarios [10], ubiquitous computing environments were they were tracked using RF and IrDA [20], moving information between computing platforms [19]. The EcoRaft Project uses embodied mobile agents to add to the “liveness” of the presentation, which it does by allowing the types of temporal change mentioned earlier to be initiated, and to some degree controlled, by participants carrying agents in Tablet PCs. This “helps strengthen the sense of direct manipulation that the interface conveys, thus making the interface more satisfying.” [24] For the interaction to be more satisfying is for it to be better suited for learning. [1]
Didactic games like Geney [4] have used such agents, but none have used 3-Dimensional embodiments or virtual environments as the EcoRaft Project does. The added dimensionality reduces cognitive load by reinforcing the illusion that the virtual spaces are extensions of the real space.
Implementation
The EcoRaft Project
The EcoRaft Project is an extension to the Virtual Raft Project. [26] The installation consists of three stationary computers that represent virtual islands, and three mobile devices that represent rafts or boxes. Each virtual island represents a different ecosystem. The ecosystems can be populated with hummingbirds, coral trees, and heliconia plants. Participants use the mobile devices to transport species from one island to another by bringing a mobile device near one of the stationary computers, allowing an animal or seed to leap onto the raft, and allowing the cargo to disembark into another stationary computer. One of the virtual islands acts as a national forest, a fully populated ecosystem that can act as a reserve, while the other two virtual islands can be deforested by the press of a button. Participants can repopulate a deforested island by bringing different species to it in a proper order by means of the mobile devices. It is easy to destroy a habitat, but although difficult to revitalize, by observing some important rules of ecological succession, it can be done. [2]
In general, a system that employs embodied mobile agents requires more than one computing platform that supports a software framework in which the agents can operate. While the physical platforms can be homogeneous, a heterogeneous mix of mobile and stationary platforms with some form of inter-device communication, such as TCP/IP, provides a richer engagement space for interacting with agents. Heterogeneous, accommodating, and virtual platforms enrich this engagement further.
The hardware for the EcoRaft Project includes three desktop PCs with onboard sound and video cards, and three tablet PCs with accelerometers. The desktop computers act as virtual islands to host embodied mobile agents, and the tablet PCs represent rafts that allow the agents to travel between the islands. The current implementation allows embodied agents like the EcoRaft Project’s hummingbirds and seeds to move from island to raft and vice versa. The computers currently run similar Java code bases, and while the hardware platforms vary and are of off-the-shelf variety, the virtual platforms are homogeneously provided by the Java Virtual Machine (JVM).
Several techniques were used to facilitate the movement of embodied agents between platforms in an engaging manner. Although the general principles behind the cross-platform motility of agents have been vetted elsewhere [10, 19, 20, 26], this project provides an interesting combination of techniques to create a richly engaging interaction between agents and people.
CTML
The EcoRaft Project was developed in line with Mayer’s seven principles for the design of multimedia messages. These principles are:
1. Multimedia Principle: Students learn better from words and pictures than from words alone.
2. Spatial Contiguity Principle: Students learn better when corresponding words and pictures are presented near rather than far from each other on the page or screen.
3. Temporal Contiguity Principle: Students earn better when corresponding words and pictures are presented simultaneously rather than successively.
4. Coherence Principle: Students learn better when extraneous words, pictures, and sounds are excluded rather than included.
5. Modality Principle: Students learn better from animation and narration than from on-screen text.
6. Redundancy Principle: Students learn better from animation and narration than from narration, animation, and on-screen text.
7. Individual Differences Principle: Design effects are stronger for low-knowledge learners than for high-knowledge learners and for high-spatial learners than for low-spatial learners. [13]
EcoRaft is certainly primarily an interactive graphical project, satisfying the first principle. Textual prompts and rewards are seamlessly embedded in the scenes they relate to, satisfying the second and third. Only informative material pertinent to narration, the agents, and the environment exist, satisfying the fourth. The narration is insufficient in noisy environments, so text had to be added. Such as it is, the text doesn’t contribute to a split-attention effect because the box it is in animates as it reveals new information. This draws a participant’s away attention from less important animation events and to more pressing information on changes in the state of the environment. Thus, the modality and redundancy principles are accommodated. The project was designed for high-spatial learners in mind, and by choosing the target age group of children 8 to12 years old, it ensures that the seventh principle is also adhered to.
General Design Considerations
To avoid obfuscation and to include as much complexity as the design team felt necessary to keep the project engaging, layering and separation were used in the graphics and sound. As Tufte asserts, “What matters—inevitably, unrelentingly—is the proper relationship among informational layers. These visual relationships must be in relevant proportion and in harmony to the substance of the ideas, evidence, and data conveyed.” [28] Hummingbirds and seeds were designed so that their details emerge as they move toward participants and the significant plants always stand out in relief against a blurred background of muted greens depicting other flora in the rest of the rainforest. This not only reduces computational expense in terms of calculating the lighting effects on the animated agents, but also emphasizes the important objects in the display. This effort complements Imhof’s first rule: “color spots against a light gray or muted field highlight and italicize data, and also help to weave an overall harmony.” [28] The high contrast also lowers the overall perceptual complexity of the presentation, and reduces the need to visually search for salient features. Similar tactics were used in designing the virtual rafts/boxes.
Several more principles on designing animation, cited by Thomas, were also exploited:
1. Characters and objects should seem solid.
2. Exaggerating the behavior of user interface objects makes the user interface more engaging.
3. The interface should reinforce the illusion of reality. [24]
Characters in EcoRaft were designed to appear solid by virtue of the way they interact with each other and the boxes. In a box, a seed will roll with the tilt of the tablet PC and make a sound as it caroms off sides it hits. Hummingbirds interact with the flowers on the plants and react to the tilting of a box as if confined to it, constantly struggling to get to the highest point.
Many things were exaggerated in the project to make it more engaging. Seeds drop out of trees and fly at the screen toward a nearby box holder, the colors of the plants are brighter than in reality, the hummingbirds are far more strikingly iridescent than they actually are, and the sounds associated with all the agents are hyperrealistically musical.
The illusion of reality is maintained by the smoothness of the translations of agents from platform to platform, the continuity of agents’ behavior in different environments, and the presences of tablet PCs as boxes for the agents, which allows participants to control and so better integrate virtual material with real space.
While these principles are not considered in Mayer’s work, they certainly complement it: without some measure of engagement, participants can’t be expected to pay attention to presented material.
Sound
Sound, as an integral part of movies, is a well explored topic. [25] Music often provides a sense of continuity between scenes. Similarly, when used as transition indicators, sounds in a mobile agent framework convey a more convincing sense of continuity between real and virtual spaces than animation alone is able to as agents move between platforms. For instance, on an island in the EcoRaft Project, a seed bounces on its way to a nearby raft after it is generated by a plant and emits a sound like a slide whistle playing an ascending scale. The seed then appears to roll into the raft from the direction of the island, accompanied by the same slide whistle playing a descending scale. This adds to the illusion that the seed is moving out of the screen. Similarly, transfer between rafts and islands is by hummingbirds is accompanied by the crescendo and diminuendo of a zipping sound characteristic of real hummingbirds. Sounds played in reverse order are used when the order of the platforms departed and entered are reversed.
Using conjunctive sounds emitted from the mobile and stationary devices during the transfer of agents supports the illusion that when a single agent leaves one platform the same agent then appears on the other. But, sound is also used to signal transitions in an island’s ecosystem. When an island’s ecosystem is deforested, ambient sound vanishes to reappear in increasing complexity as more flora and fauna are added to it. An additional musical score associated with the hummingbirds also textures the soundscape and adds to their overall attractiveness.
Using sound in these ways strengthens the apparent connection between the embodied agents and their surroundings by giving the virtual and real spaces complementary contributions to the overall acoustic environment.
Graphics
Agents should have a striking appearance that contrasts enough with their surrounding milieu so as to be easily tracked by an observer as they move about on any platform. Hummingbirds are naturally iridescent and both Heliconia plants and coral trees have lush and vibrant flowers. But, in the tropical landscape of a rainforest, all of these might be missed without some enhancements. Lighting, color, dimensionality, and movement are used to set these apart from their backdrop in the EcoRaft Project. In this regard, believability, more than reality, is an important aspect of being engaging. [15] By slightly exaggerating the look of key agents, in giving them richer textures and modeling them in 3-D, and by diminishing the visual impact of their background (created through the use of filters and 2-D images), the stage is set for a more dynamic interaction between agents and users. For instance, although seeds don’t normally bounce a great distance from a plant to land in a box accompanied by a whistling sound, this behavior is acceptable in The EcoRaft project; other aspects of the project are just hyperreal enough to allow for this without caricaturizing the serious goal of the project (to teach concepts of ecological restoration.) Thus, the look and behavior of the agents should complement their environment and maintain the metaphor of the simulation. [6]
An additional consideration is frame rate. As the frame rate of an animated scene dips below 30 fps, so does the attraction of the scene and its characters. To prevent this in the EcoRaft Project, level of detail techniques were employed.
Behavior
In the EcoRaft Project, complex behaviors are exhibited primarily by hummingbirds, which currently can feed, perch, or starve, or fly between stationary computers and mobile devices. As mentioned earlier, the behavior of agents should complement the ambience of their environment. Yet, this is hardly enough in itself to make an agent engaging. To achieve a higher level of engagement, embodied agents must act and interact in a manner that entails a sufficient amount of complexity found in correlative real system. When the behavior is suitably complex, a participant gets the impression of observing a believable and coherent system, while experiencing activities s/he may have not been privy to otherwise. An acceptable illusion of reality is maintained although the artificial system does not have the full complexity of the real one.
Embodied agents can be further enhanced if allowed to interact with their observers directly. In the EcoRaft Project, a webcam is used to detect motion in front of an island. This signals one of the plants on the island, a centropogon vine, to sway, and increases the likelihood that existing hummingbirds will fly toward participants and display themselves. In doing so, the birds usually look forward, appearing to be enrapt momentarily by any observer before flying off.
Real-world sensing and interactivity like this lends credibility to an agent. The outward gaze of the hummingbirds at nearby people is designed to enhance the sense of continuity between the virtual world of the hummingbirds and real space and further engage our attention. [1]
Compelling Interaction
These behaviors and interactions encourage a sense, within the user, of participation in the agent’s behavior, and thus add to the quality of the engagement. Allowing 3-D embodied agents to move through networks of heterogeneous stationary and mobile devices as a result of human action fosters this sense of connectivity. Furthermore, by allowing an agent to explore the confines of the visible virtual space of its mobile and stationary devices adds naturalness to its behavior. Of course, a seed wouldn’t fly about the virtual confines of a raft on its own, but with the addition of accelerometers to the platform, the seed can roll about in response to a user tilting the device in a realistic fashion. Similarly, a hummingbird might be expected to fly to the highest part of its particular confines as that device is tilted.
Part of the novel experience of the EcoRaft Project is the apparently seamless way plants and hummingbirds transfer from stationary computers to mobile devices and vice versa. The smoothness of the transition helps minimize behaviors unexpected by participants, and helps maintain the illusion of continuity between the real and virtual spaces, thus eliminating some undesirable complexity from the environment of the project.
Evaluation
Thus far, the EcoRaft Project has been presented at the SIGGRAPH 2005 Emerging Technologies exhibit, at several other conferences, twice at the Discovery Science Center in Orange County California, and in a research lab at UCI. During the presentations, semi-formal investigations were conducted into how engaging EcoRaft is and how effective it is as a pedagogical tool. Questionnaires were used, and participants taken aside and interviewed. Additionally, some of the design team took notes while observing participants interacting with the installation. The overall attitude of the participants to the project was enthusiasm, with few complaining that it was too simplistic or too complex. Several claimed to have learned something about restoration ecology, but no research was conducted to investigate whether such knowledge was going to be retained long-term or whether it was transferable.
What was evident was that peer-to-peer instruction was occurring as some children aided others, and accompanying adults offered the children some instructional scaffolding. The EcoRaft Project was offering a space for learning to be socially constructed.
Discussion
Complexity in the project was reduced by introducing clear cues as to what information should be attended to at any particular time. EcoRaft was specifically designed to deal with the putatively multimodal way children of ages 8 to 12 resemiotize data. The appropriateness of the exhibit for this demographic was validated when, from both observations and questions made by the design team, it was apparent that younger children enjoyed playing with mobile agents but didn’t grasp the overall theme the way the children in the intended age group did.
Unfortunately, the questions asked by the design team did not seek to reveal whether any knowledge constructed by participants was meaningful. Meaningful knowledge is that which can be retained and transferred. [13] Although, with the narrative guidance of construction provided by other participants, the design team, and the project itself, it may be expected that the children engaged in an active process of meaning making. [18] This implies that some retention and transferability of information did occur.
To increase the significance of analyzable data, both retention and transfer questions should be asked of participants. Retention questions should be explanative, that is ask how to perform certain tasks. Transfer questions should include inquiries about redesign, troubleshooting, prediction (inference), and conceptualization. [13] Below is a list of some suggested questions in respective order.
Retention
Explanation: What is the order of succession that must be used to repopulate an island?
Transfer
Redesign: How would you modify the project to repopulate an island faster?
Troubleshooting: If a Heliconia seed wasn’t germinating, how would you make it grow?
Prediction: If the hummingbirds vanished from the rainforests of Costa Rica, what would happen to the trees?
Conceptual: Why are hummingbirds important to rainforests?
Scoring the answers to these questions like these would also allow metrics to be collected and analyzed. This would better reflect the pedagogical success of the EcoRaft Project.
While generally following Mayer’s guidelines, the Project also uses other principles of artistic design and animation. These are necessarily employed and in a complementary fashion because of the need to incorporate 2-D narrative text in the 3-D environment and because the design team hoped to exploit the impact of levels of engagement on the learning experience of participants. The studies of multimedia learning have mostly focused on static text. The design team has observed that animation and narration that guide the gaze of onlookers toward informative text associated with a multimedia presentation enhances their overall experience. This might have worked out differently had the exhibit required more attention to the details of the activity in the virtual environments, including speedier responses on the part of the participants, but this was alleviated by the overall implemented considerations of balancing simplicity and complexity. Mayer’s theories can be extended with the other principles employed in EcoRaft.
Future Work
In the future, the question-asking strategy will provide more salient data, and the exhibit will be shown to other audiences. Current development seems driven towards a Flash version of the EcoRaft project, which should remove some of the cognitive load of participants through the use of simpler, vector graphics. This shift will thus also allow the developers to include more complex agent behaviors while incurring less computational overhead. Ultimately, it will make the project more accessible to potential users through the Internet, giving children greater access to EcoRaft’s important message.
Conclusion
The EcoRaft Project extends the principles of Mayer’s CTML by incorporating principles of artistic design, by combining 2- and 3-dimensional effects, and by further exploiting the benefits of multimodal learning through encouraging haptic and social engagement. In doing so, the project provides a novel paradigm for the presentation of multimedia in a pedagogical setting. With the introduction of metrics to the project, the design team may be able to generalize the techniques it used in a way that exceeds the usability of CTML alone.
References
1. Bickmore, T.W. and Picard, R.W. Establishing and maintaining long-term human-computer relationships. ACM Transactions on Human-Computer Interaction (TOCHI), 12 (2). 293-327.
2. Carpenter, F.L., Nichols, J.D., Pratt, R.T. and Young, K.C. Methods of facilitating reforestation of tropical degraded land with the native timber tree, Terminalia amazonia. Forest Ecology and Management, 202. 281-291.
3. Cooper, G. Research into cognitive load theory and instructional design at UNSW, School of Education Studies, The University of New South Wales, Sydney, NSW 2052, Australia, 1998.
4. Danesh, A., Inkpen, K., Lau, F., Shu, K., and Booth, K., GeneyTM: designing a collaborative activity for the palmTM handheld computer. in Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI '01), (Seattle, Washington, United States, 2001), ACM Press, New York, NY, 388-395.
5. Dunne, C.R. Using Mobile Agents for Network Resource Discovery in Peer-to-Peer Networks. ACM SIGecom Exchanges, 2 (3). 1-9.
6. El Refaie, E. Understanding visual metaphor: the example of newspaper cartoons. Visual Communication, 2 (1). 75-95.
7. Hofmann, M.O., McGovern, A. and Whitebread, K.R., Mobile Agents on the Digital Battlefield. In Second International Conference on Autonomous Agents. in Second International Conference on Autonomous Agents (Agents '98), (Minneapolis, Minnesota, 1998), 219-225.
8. Iedema, R. Multimodality, Resemiotization: Extending the Analysis of Discourse as Multi-semiotic Practice. Visual Communication, 2 (1). 29-57.
9. Jansen, W. and Karygiannis, T. Mobile Agent Security. NIST Special Publication 800-19.
10. Kangas, K.J. and Roning, J. Using mobile code to create ubiquitous augmented reality. Wireless Networks, 8 (2/3). 199-211.
11. Large, A., Beheshti, J., Breuleux, A. and Renaud, A., The influence of multimedia on learning: a cognitive study. in Proceedings of the Second ACM international Conference on Multimedia (MULTIMEDIA '94), (San Francisco, California, 1994), ACM Press, New York, NY, 315-319.
12. Lowe, R.K. Animation and learning: Selective processing of information in dynamic graphics. Learning and Instruction, 13. 247-262.
13. Mayer, R.E. Multimedia Learning. Cambridge University Press, 2001.
14. Ormerod, F. and Ivanic, R. Materiality in children's meaning-making practices. Visual Communication, 1 (1). 65 - 91.
15. O'Sullivan, C., Dingliana, J., Giang, T. and Kaiser, M.K. Evaluating the visual fidelity of physically based animations. ACM Transactions on Graphics (TOG), 22 (3). 527-536.
16. Page, J., Zaslavshy, A. and Indrawan, M. A buddy model of security for mobile agent communities operating in pervasive scenarios. ACM International Conference Proceeding Series, 54. 17-25.
17. Palmiter, S. and Elkerton, J., An evaluation of animated demonstrations of learning computer-based tasks. in SIGCHI Conference on Human Factors in Computing Systems: Reaching Through Technology (CHI '91), (New Orleans, Louisiana, 1991), ACM Press, New York, NY, 257-263.
18. Plowman, L., Luckin, R., Laurillard, D., Stratfold, M. and Taylor, J., Designing multimedia for learning: narrative guidance and narrative construction. in SIGCHI Conference on Human Factors in Computing Systems: the CHI Is the Limit (CHI '99), (Pittsburgh, Pennsylvania, 1999), ACM Press, New York, NY, 310-317.
19. Rekimoto, J., Pick-and-drop: a direct manipulation technique for multiple computer environments. in 10th Annual ACM Symposium on User interface Software and Technology (UIST '97), (Banff, Alberta, Canada, 1997), ACM Press, New York, NY, 31-39.
20. Satoh, I. SpatialAgents: integrating user mobility and program mobility in ubiquitous computing environments. Wireless Communications and Mobile Computing, 3 (4). 411-423.
21. Slobodkin, L.B. Simplicity & Games of the Intellect. in Simplicity & Games of the Intellect, Harvard University Press, 1992, 8.
22. Sweller, J., van Merrienboer, J.J.G. and Paas, F.G.W.C. Cognitive architecture and instructional design. Educational Psychology Review, 10. 251-296.
23. Tabbers, H.K., Martens, R.L. and Van Merriënboer, J.J.G. Multimedia instructions and cognitive load theory: Effects of modality and cueing. British Journal of Educational Psychology, 74. 71-81.
24. Thomas, B.H. and Calder, P. Applying cartoon animation techniques to graphical user interfaces. ACM Transactions on Computer-Human Interaction (TOCHI), 8 (3). 198-222.
25. Thompson, B. Evoking Terror in Film Scores M/C: A Journal of Media and Culture 5, no.1, M/C, 2002.
26. Tomlinson, B., Yau, M.L., O'Connell, J., Williams, K. and Yamaoka, S., The virtual raft project: a mobile interface for interacting with communities of autonomous characters. in CHI '05 Extended Abstracts on Human Factors in Computing Systems, (Portland, OR, 2005), ACM Press, New York, NY, 1150-1151.
27. Tudoreanu, M.E., Designing effective program visualization tools for reducing user's cognitive effort. in Proceedings of the 2003 ACM Symposium on Software Visualization (SoftVis '03), (San Diego, California, 2003), ACM Press, New York, NY, 105-ff.
28. Tufte, E.R. Envisioning Information. Graphics Press, 1990.
29. Tuovinen, J.E., Optimising student cognitive load in computer education. in Australasian Conference on Computing Education (ACSE '00), (Melbourne, Australia, 2000), ACM Press, New York, NY, 235-241.
30. Westelinck, K.D., Valcke, M., De Craene, B. and Kirschner, P. Multimedia learning in social sciences: limitations of external graphical representations. Computers in Human Behavior, 21 (1). 555-573.