What pedagogic approaches were used?

The researchers began their development efforts with the observation that “one of the key properties of virtual environments is their ability to captivate.” (Scaife, 2001) Using a theoretical model the researchers called “external cognition,” the environment supported both individual engagement within the virtual theatre and reflective thought outside it to “offload cognitive effort” into an external representation such as writing and editing, as well as discussion among children. (Marshall, 2002)

This environment was designed for pre-literate children as a tool to promote narrative play as well as reflection about that play. The goal was not direct acquisition of knowledge, but an extending of learners’ imaginations. The learners constructed stories in the virtual theatre to help them develop their own sense of narrative. The assumption of this approach was that it would contribute both to their learners’ creativity and future literacy. The goal, as the developers described it, was “to provide young children with a means of extending their existing repertoire of story telling by providing them with a new set of tools they could use to create, edit, direct, and act out plays in a virtual, imaginary setting.” (Scaife, 2001)

One of the features of this design approach is that it allowed a “stepping out” of the story the child is immersed in to reflect on the story as well as view the story from different characters’ points of view, promoting their understanding of different roles in story development and enactment.

The design of the environment was greatly affected by the fact that its users could not read. All commands in the interface and all directions for use had to be presented visually. This posed a significant problem for the designers, but later tests with learners of this age indicated that they succeeded. Research was also done in the early stages of the design to understand the competencies of children of this age group around interacting with both physical and virtual story settings to inform the designers of their capacities as storytellers and manipulators of plot, character, and action.

The testing of the product was done in classrooms of children ages 4-6 and 7-8, usually in pairs. The developers envisioned the environment being used by one or two children at a time, but not more. Teachers and adults were involved in the environment to help the child understand how to interact and to facilitate story development, but otherwise, no “greater community” was envisioned.

The results of the design studies indicated that their pre-literate learners could understand the different interaction styles and switch between them. For example, they could choose between different avatars and could understand the different point of view that avatar had. Their learners did not need the forms of representation for characters, animals, and barnyard to be elaborate. In fact, the visually rich environments that they tested were more distracting than inspiring. Research on the design found that the more that is left to the children’s imaginations, the better they responded creatively.

The children were able to create stories on their own without parts of the story given to them. Their ability to determine a beginning, middle, and end to the story was enhanced by switching from first person “in the play” mode to the third person writing and editing mode. This mode switching also helped the children make the characters behave differently to further the story, and then go back into the story and enact different experiences.

Reviewer’s Note:

As our ability to send and receive data over the Internet increases, virtual learning environments will become more media rich and interactive. Soon we will have more 3D environments like PUPPET in which our children can learn. How shall that wealth be used? The PUPPET Virtual Theatre gives us an excellent example of not only of what is possible, but what is valuable. The value of the PUPPET environment is two-fold (at least). It prepares children for literacy, a challenge in which so many children falter. Secondly, it develops creativity, a rare accomplishment in any learning environment.. In many ways, PUPPET is an example of what may be the highest and best use of technology in the service of learning: to individually prepare children to succeed in school by preparing their minds for the tasks ahead. There is a Harvard-based game development company that is doing something similar with children who have learning disabilities. The “get kids addicted to working hard, thinking hard, and thinking complexly…to develop the habit of creative independent thinking.” (Brittan, 2004)

PUPPET is all about making stories, developing characters, verbalizing emotion, learning to see the world from different points of view, each of which requires creative thinking. The kind of creative thinking is different than the “creative problem-solving” of Jasper where learners work to achieve a goal (the problem solution). PUPPET has no goal except the expression of ideas, the use of the mind. One of the most exciting features of virtual environments is how they allow the physically impossible to happen virtually, giving wide range to creative activity. That means that children can have virtual experiences unlike anything ever done in a classroom. PUPPET serves to remind us to occasionally stop thinking of “scope and sequence” and do more thinking of human development.
The GenScope Learning Environment

What are descriptors of the project?

Research group: Genscope Development Team had members from the Concord Consortium, various universities such as the University of Georgia, and the Educational Testing Service.

Learner age group: Secondary school science learners of multiple ability levels (basic science to honors).

Technical description: GenScope software initially ran on 33 Mhz Macintosh computers from the early 1990s in computer labs, but over time Genscope was used on laptops in classrooms. Genscope presented different aspects of genetics information in graphic representations in multiple windows, coordinated with paper-based activities and assessments. GenScope was described as an “open-ended exploratory software tool.” (Hickey, 2003, p. 495)

Time in development and use: Software development began in 1991, research collaboration began in 1995 between developers and the assessment team. Research reported in 2003.

How did the project evolve over time?

The GenScope project had three phases during which the environment changed from one that was rich in genetics exploratory activities to one that was more aligned with its summative assessment. During this evolution, the researchers “narrowed” the scope and refined the activities to include more formative assessment to better prepare the learners for the summative assessments that had been developed based on national standards for genetics content. The phases were: Year 1: piloting, revisions and formative assessments, Years 2 and 3: large-scale implementation and evaluation, and Year 4: follow-up study.

The primary goal was to help learners develop a cognitive model of genetics based on the way experts in genetics solved genetics problems. Careful effort was made to align the learning with the assessment of that learning. To achieve this, during its evolution, the environment added more formative assessments to increase transfer from the GenScope environment to the assessment environment because Phase 1 research showed that learners had not gained a meaningful domain understanding in genetics. Once this was done, the entire environment worked well and summative assessments showed a significant gain in genetics domain knowledge and problem-solving abilities.

Most activities involved dragons. Learners used the windows in which key information about the organism, its DNA, its pedigree, and its chromosomes were displayed. Additional windows showed the meiosis process and the population showing the genetic trait in question, such as horns and wings. Learners worked back and forth in the various windows to gain knowledge about the dragons’ traits and heredity to solve problems posed by the paper-based activities.

The learners could construct their own knowledge by changing DNA order, choosing what genetic traits to look at, and causing the system to randomly trigger mutant traits such as albinism and double wings. When the initial open-ended exploratory environment yielded assessments that indicated the design had not focused sufficient attention on genetics concepts that could transfer to an assessment environment, the researchers used a situative analysis of transfer to “compare the resources that support meaningful participation in the learning environment.”

What key design features were developed to support the creation of a community of learners?

In the large scale implementation (Phase 2), there were 31 classes taught by 13 teachers. Many of the teachers were recruited for the study because they had independently downloaded the software from the GenScope website. There is no indication in the research of any attempt made to create a community of learners in any of the phases, but the primary report of this environment focused on its development and assessment alignment and did not go into detail about implementation, so the community aspect may simply have not been reported.

In each phase, learners were assessed and the curriculum was revised to improve it in the areas shown as weak by the assessments. By the time the follow-up study was done, there were laptops in the classrooms and revised “Dragon Investigations” that aided transfer from the colorful interactive environment of GenScope to the paper-based assessments that were more like what learners would later encounter in large-scale testing. The results were truly impressive. The combined gains of the GenScope classes were significantly above the gains of a comparison class which followed the traditional genetics teaching approach with a textbook. The scores showed gains of 13.3 for the comparison (traditionally taught) class and 22.6 and 30.7 for the GenScope classes, with the lower scoring class using only the GenScope software but not the “Dragon Investigations” whereas the higher scoring class used the complete curriculum. The researchers concluded: “We believe it [the results reported above] demonstrates the potentially dramatic knowledge gains possible when teacher knowledge, curriculum, technology, classroom assessment, and external assessment are aligned towards well-defined, ambitious goals.” (Hickey, 2003 p. 527.

Research on the GenScope environment was included because it illuminates so many issues. First is the issue of “teaching to the test.” The developers readily admit that they “narrowed” the rich field of exploration they had built to focus their learners and thereby assure better test scores for them. Obviously, this issue gets to the very heart of the role of education itself. Perhaps one should look at the fact that this issue is appearing in a virtual learning environment as proof that they are “coming of age,” growing up enough to have the same problems as other realms of education. Secondly, in GenScope, it is important to note the surprising treatment of content: the use of dragons as the objects of study, having wings and horns appear as dominant and recessive traits, and having gross mutation cause unicorns. For years, genetics was taught with barely visible fruit flies as its exemplar. I commend the designers of this environment for using the “non-real” aspects of virtual worlds to full advantage. The result was a more fun, more creative treatment of genetics that engaged learners in the subject. Their results showed that the traditionally least engaged types of science students, those in basic or vocational science classes, showed the most gain when assessed in their genetics knowledge (scores comparing these GenScope students to their non-GenScope classmates were 30.7 and 13.3 respectively). Genetics is such a visual science, with its double helix DNA models and its observational traits, that it was a good choice for a virtual learning environment. The results showed that they succeeded in engaging their learners. For those two reasons, the GenScope research was an important addition to understanding what’s really going on in the practice of virtual learning environments.

The project began with the goal of “making learning fun,” but after 30 months of ethnographic research, they changed the focus of the design to include a broader social commitment. It is key to understanding this and future virtual learning environments to know that it was the children themselves who wanted more than fun; they wanted to be engaged in personal, social, ethical, and environmental issues. (Barab, 2005 p. 87)

Learners completed “quests” that were connected to the narrative of Atlantis being a world in trouble. The storyline did not reside in one location or medium, but came together as the user participated in the game context and investigated “relevant personal issues.” (Barab, 2005, p.87)

Quest Atlantis built on strategies from online role-playing games to engage users and blended them with constructivist, situated cognition theories of learning which focus on the centrality of activity. They used Vygotsky’s “novel stance toward play” citing his opinion about the role of play in learning as “…imagination is adolescents and school children is play without action.” (Vygotsky, 1933/1978)

As part of their approach, the developers intentionally brought in media and game designers to increase the level of engagement of the environment because they felt that these designers were “most successful in engaging children.”

The learner worked with all of the artifacts plus the environment itself to construct an answer, or a series of answers, to solve a quest. The environment had a participatory framework including hands-on action as well as reflection. The general approach to the solution of the quests was “inquiry-based learning” and was central to the design of the environment. (Barab, 2001) In addition to inquiry, learners were asked to contribute their expertise and ideas to the environment, as well as engage in social issues that have local relevance and “report in” those activities to blend the Atlantian and real worlds.

The developers used the theory of the zone of proximal development as it applies to play. They designed the environment to support unrestricted, free activity that encourages children to behave beyond their ages. It also liberated them from many of the social constraints of real-world play and other forms of social activity.

The use of the principles of formative assessment were integrated with the learning process by having learners create portfolios of their work that could later be assessed by teachers and other members of their learning community.

This environment had a huge emphasis on community; it even involved its learners in their own communities. Quest Atlantis was designed with many interactive opportunities based on the designs of “persistent virtual worlds,” which are universes with their own cultures and discourses which they have adapted to engage children in learning. The learners in this environment learned from each other, from teachers, from experts distributed across the environment, and from community members they engaged with on their quests.

In their studies on the impact of QA on learning, the researchers found that the learners offered character insights that were “deeper or better supported” than did learners in equivalent conditions, and they demonstrated statistically significant learning over time in the areas of science, social studies, and sense of academic efficacy. (Barab, 2005) They also queried the teachers using Quest Atlantis and those teachers said they chose it because of:

• 
  Social commitments
  
• 
  Direct connection to academic standards
  
• 
  Use of technology is engaging to student (This is shown in the fact that children in schools and after-school settings have completed hundreds of quests without any mandated requirement.)
  

The latest research report about Quest Atlantis (Barab, 2005) was entitled Making Learning Fun: Quest Atlantis, a Game Without Guns, yet the reason Quest Atlantis is in this review is because its 30 months of qualitative research showed that children wanted “more than fun” and the designers evolved the environment in response to them. The children in that study wanted to make the world a better place, they wanted to learn the skills to do that, they wanted to work together with friends to get started, and they wanted to be part of a community for action. Whew! It makes fun look trivial, doesn’t it? Of course, what is being revealed in this data is that the real fun of life, as children told the Quest Atlantis researchers, is being involved, living with purpose, and working toward a lofty goal. (Not coincidentally, these are the same beliefs that the leaders of the human potential movement say work with adults who are seeking happiness in their lives. (Leider & Shapiro, 2002))

Quest Atlantis may have started out trying to “make learning fun” but it has evolved to so much more. The fact that Barab is now talking about “sociocultural design,” and has created an environment that he describes as “at the intersection of education, entertainment, and social commitment” (Barab, 2005) is groundbreaking.

One of the themes of this review has been that virtual learning environments can offer learner experiences never before available in education. Quest Atlantis may be the bellweather of the rich, multifaceted offering of virtual learning environments of the future, taking children farther than we can perhaps image.

Summary: Common Trends, One Notable Difference

When the theoretical approaches to design are compared, it is clear that four “big ideas” are very much in use: the first is the repeated presence of constructivism, having learners construct their own knowledge within the environment instead of having it presented to them to absorb passively. The second “big idea” is the awareness that learning is situated and, because of that, there may need to be more conscious efforts to support transfer of knowledge from these sometimes media-rich 3D colorful environments to other demands for that knowledge, such as paper and pencil assessments. The third “big idea” that was present, although sometimes obscured, was the importance of creating multiple zones of proximal development for learners in the environment. The design challenge this need presents is significant, but solvable, especially if it is a design priority.

The final “big idea” present in the research about virtual learning environments was the practice of expanding the environment to embrace a community of learners. There was a consensus that the advantages of creating a community of learners as part of the design of an environment added tremendously to the value of the environment for all its learners. We are truly raising a generation of “connected” children who share their knowledge and work in teams and the community of learners idea will grow stronger as these virtual worlds expand. The only environment that didn’t attempt to create a community was the PUPPET Virtual Theatre. One reason for this was that the virtual theatre was a highly immersive environment designed to encourage the creativity of a pre-literate child. This environment may show that there are, indeed, some environments that should not be designed to be expandable to community involvement. It may be that certain creative efforts are best designed to be solitary. As the PUPPET virtual theatre evolves, which it will, it will be interesting to see if the children who use it signal a desire for more interaction with other learners while working in their virtual theatres.
Some Thoughts about the Future
This review indicates that how we design experiences that truly allow children to construct knowledge is a doubly-difficult problem in virtual reality. Missing is the “guide” needed for “guided discovery” in many cases, yet present is the ability to give constant feedback, which if used to guide efforts to construct knowledge and solve problems could replace this deficiency somewhat. Not present in these virtual environments is a direct “mentoring” aspect that could serve more of the functions of the teacher if an environment evolved from being blended into a classroom to being totally virtual. This seems like a logical evolution, done so that virtual environments could be used by more learners in different situations; however, the design efforts to do so need to be consciously focused in that direction.

The question of transfer from virtual reality to real school and real life was in need of more research: just what is learned online can transfer, and how does that transfer affect schooling success and life skills? Can this be done without the “narrowing” of the environment, like what that needed to happen in the GenScope environment to yield good assessment results? (Hickey, 2003)

How a virtual learning environment can assess learners and respond to them by providing them with multiple zones of proximal development was uncertain at best, but probably the area of greatest gain if solved. This is especially critical to investigate because classrooms are becoming more diverse. For example, a class may have an increasingly wide range of learners’ reading abilities, especially if the class has second language learners. This diversity of reading levels makes additional demands on the teacher, who is already overburdened. A virtual learning environment that can assess a learner’s reading level and challenge him or her at that level would attend to the needs of that student, and every student in the class without making additional demands on the teacher. With more attention to this important design feature, it could be possible for virtual learning environments to teach to all levels of learners in a virtual community, whether that learner’s classroom can do that or not.

How the term “community of learners” was used requires further investigation. “Communication among learners” is not necessarily the same as a “community of learners.” This point could easily be missed by future developers of virtual learning environments, who may think that providing opportunities for communication is the same as causing a community of learners to happen. Community requires learners seeking their own places in a community, and understanding the contribution they personally make to contribute to that community’s learning goals, to have a true “community of learners.”

References

Cognition and Technology Group at Vanderbilt, (1997). The Jasper Project: Lessons in curriculum, instruction, assessment, and professional development. Mahwah, N.J.: Lawrence Erlbaum.

Pea, R.D., (1992). Distributed multimedia learning environments: The CoVis project. Commun. ACM (May 1992), p. 60-63.

Scaife, M. & Rogers, Y. (2002). Informing the design of a virtual environment to support learning in children. International Journal of Human-Computer Studies, 55. pp. 115-143.