Games for Research Purposes: The Case of Crystal Island
Beck and Wade’s (2004) characteristics are also evident in non-commercial games, such as Dede’s long standing River City, an immersive simulation for middle school students. This MUVE (Multiuser Virtual Environment) is an example of an academic enterprise that was created using designed-based research and promotes both complex communication and expert problem solving. Following the path of River City, with the addition of intelligent tutors, Crystal Island, is being developed at North Carolina State University by a team of computer scientists and educational researchers. This NSF funded project is an example of an academic innovation that targets science education for eighth grade middle students. Taking their cues from Jerome Bruner (1990), who observed that the way people organize their experience and knowledge with the social world “is narrative rather than conceptual” (p.35), the creators are using a narrative centered learning environment to explore concepts related to microbiology (Mott & Lester, 2006). The learning environment (see Figure 1) is set on a recently discovered volcanic island where a research station has been established to study the unique flora and fauna. The user plays the role of the daughter (or son) of a visiting scientist who is attempting to discover the origins of an unidentified illness at the research station. The environment begins by introducing the student to the island and the members of the research team for which her father serves as the lead scientist. As members of the research team fall ill, it is her task to discover the cause of the outbreak. She is free to explore the world to collect physical evidence and interact with other characters. Through the course of her adventure she must gather enough evidence to correctly choose among candidate diagnoses including botulism, cholera, salmonellosis, and tick paralysis, as well as identify the source of the disease relying on her knowledge of genetics to solve the mystery.
(a) (b)
Figure 1. The CRYSTAL ISLAND learning environment.
The task-oriented environment of Crystal Island, its semiautonomous characters, and the user interface are implemented with Valve Software’s Source™ engine, the 3-D game platform for Half-Life 2. The user can perform a broad range of actions including performing experiments in the laboratory, interacting with other characters, reading “virtual books” to obtain background information on diseases, and collecting data about the food recently eaten by the members of the research team. Throughout the mystery, users can walk around the island and visit the infirmary, the lab, the dining hall, and the living quarters of each member of the team. In the current test bed, there are 20 goals users can achieve, three hundred unique actions the user can carry out, and over fifty unique locations in which the actions can be performed.
In an attempt to increase student interest and engagement levels with Crystal Island we conducted two sets of focus studies, one with individual student sessions and the other with groups of four students. A total of 14 eighth grade students solved the mystery and then participated in conversations reflecting on their experience. Discussions centered on the topics of plot, characters, and setting; findings of these studies as they relate to interest and engagement are summarized below:
• Plot. Students enjoyed the science mystery but suggested that more conflict would add to the story. Recommendations included developing conflicts between characters and introducing misleading storylines, such as a rogue researcher accused of poisoning the sickened team members. These additional plot elements would presumably enhance students’ curiosity, supporting the goal of enhancing intrinsic motivation. Students also focused on the story’s introduction, recommending that a “training” level be utilized to establish plot, characters, and setting. Discussions emphasized an interactive scenario over a scripted cinematic sequence, providing exercises to establish navigational controls and foreshadow task objectives.
• Characters. Students suggested that Crystal Island characters needed to be more expressive and active. They wanted the learning environment to be populated by characters who could develop relationships, perhaps as active collaborators in the pursuit of solving the mystery. There was also an interest in further developing character-student conversations. Students related to other games that provide multiple-choice responses where the player has control of the dialog and perceived control of plot progression. Believable, meaningful character interactions can aid in immersing the student in the story world, thereby supporting intrinsic motivation through enriched curiosity and fantasy. Several students expressed interest in being able to design their own characters along several possible dimensions including role, gender, and physical features (e.g., hair, eyes, and clothing). This level of control over characters' appearance potentially allows students to better identify with the character, investing themselves in the story and enhancing motivation.
• Setting. During group sessions, students worked together to draw new maps of Crystal Island. The students designed the world to include areas that would evoke interest, curiosity, and opportunities for science learning. Beyond the base camp, where the current Crystal Island mystery is set, students envisioned a rich landscape replete with volcanoes, mountains, woods, caves, lakes, and streams.
Based on results of the pilot study, the research team made appropriate modifications to Crystal Island in an effort to increase interest and potential student engagement with the environment. A second data collection involved 60 eighth grade students who solved the mystery and responded to a series of questionnaires that assessed student engagement and academic dispositions (i.e., student achievement, teacher ratings of student persistence, and goal mastery orientation). We are assessing the relationship among student engagement, academic dispositions, and success in the Crystal Island environment (i.e., number of goals and actions completed and amount of time to solve the mystery). Results from this analysis, which will be presented at the 2008 American Educational Research Association meeting in New York (McQuiggan & Lester, 2008; Spires, Turner, Lester, & McQuiggan, 2008), will provide important insights into individual differences among game players that will contribute to future customization of content and game design features. Future data collections will target the effects of Crystal Island on student problem solving and affect. As a narrative centered learning environment, Crystal Island incorporates three of Beck and Wade’s (2004) gameplay characteristics: the opportunity to rapidly analyze new situations, to interact with unfamiliar characters, and to solve problems quickly and independently. Additionally, Crystal Island satisfies Malone & Leper’s (1987) criteria of challenge, curiosity, control and fantasy; learning includes competence and direction in the face of novelty, complexity, and ambiguity.
Conclusion
As educators continue the quest to ensure that all students have the opportunity to participate fully in society, multiple paths for learning must be explored. Although in its infancy, game-based technologies hold promise in forging new models of learning and teaching for the formal schooling process. Central to this challenge in the 21st century is finding cross-sector partners who are willing to take up the research and development mantel in order to shed more light on the educational benefits of games. Gee (2003) identified 36 learning elements embedded within games that he analyzed. He concluded: “Better theories of learning are embedded in the video games many children in elementary and high school play than in the schools they attend. Furthermore, the theory of learning in good video games fits better with the modern, high-tech global worlds of today’s children and teenagers live in than do the theories (and practices) of learning they see in school” (p.7). Gee’s assertion, no doubt, is designed to be provocative, but his statement holds some truth. As games become more popular, however, generalized, non-substantiated statements will not serve the field well. We need a systematic way to analyze the learning features of games and conduct educational research that will help articulate the cognitive, affective, and social benefits for education. Students are becoming increasingly vocal about what they need to be engaged and successful in school; it behooves us not only to hear what they are saying but to help create the research breakthroughs that will facilitate 21st century teaching and learning.
