The weak science achievement of U.S. elementary and secondary students reflects the uneven quality of current science education. Although young children come to school with innate curiosity and intuitive ideas about the world around them, science classes rarely foster their interest. Students spend time reading science texts, listening to lectures, carrying out preordained “cookbook” laboratory activities, and memorizing the disparate science facts that are emphasized in high-stakes science tests, increasingly losing interest in science as they move from elementary school to middle and high school.

Many experts call for a new approach to science education, based on a growing body of cognitive research indicating that science learning is a multifaceted process involving much more than recall of facts (National Research Council, 2005b, 2007, 2009). In this approach, teachers and instructional materials spark students’ interest by engaging them in exploration of natural phenomena and support their learning with several forms of instruction. Students simultaneously develop conceptual understanding of these phenomena and science process skills while maintaining their motivation for continued science learning. The new approach reflects growing understanding of the critical importance of interest and enthusiasm in scaffolding science learning.

Computer simulations and games have great potential to catalyze and support the new approach, by allowing learners to explore natural phenomena that they cannot directly observe, due to time scale (too fast or slow), size (too big or small), or form (e.g., radio waves). Learners can manipulate virtual systems that represent these natural phenomena, a process that helps them to draw powerful mental connections between the representations and the phenomena and to formulate scientifically correct explanations for the phenomena.

Overall, the evidentiary base for learning science from simulations is stronger than that for games. There is promising evidence that simulations enhance conceptual science learning and moderate evidence that they increase students’ motivation for science learning. Emerging evidence from a small number of examples suggests that well-designed games can motivate students, encourage them to identify with science and science learning, and enhance conceptual understanding—but overall the research on games remains inconclusive.

Although both simulations and games have been used for training and education for over three decades, their effectiveness for science learning has not been studied broadly or systematically. Reaching the potential of simulations and games to motivate and engage science students, enhance science achievement, and advance other science learning goals will require a stronger, more systematic approach to research and development.

The committee’s proposed research agenda outlines such an approach. The first section of the agenda focuses on improving the overall quality of the research, the second section outlines particular topics requiring further study, and the third section identifies approaches to institutionalizing research and development on games and simulations for science learning.

Research on how simulations and games support science learning has not kept pace with the rapid development of these new learning technologies. Although the evidence base related to simulations is stronger than that related to games, both areas are thin. Much research has been exploratory, making it difficult to generalize, because researchers and developers have not always clearly defined the desired learning outcomes or the mechanisms by which the simulation or game is expected to advance these outcomes.

The committee recommends that future research on simulations and games follow a design-based approach aimed at continuous improvement, including the following steps:

• Researchers and developers should clearly specify the desired learningoutcomes of a simulation or game and describe in detail how it isexpected to advance these outcomes. This should include descriptionof the design features that are hypothesized to activate learning, theintended use of these design features, and the underlying learningtheory. Researchers should also indicate direct evidence of studentlearning, if such evidence is available. This will allow research findings to accumulate, providing a base for improved designs to furtherenhance the effectiveness of games and simulations for learning.

• Researchers should initially develop methodologies for both the designand evaluation of games and simulations that focus on continualimprovement. The use of such methodologies will help to ensure thatlarge studies are not outdated by the time they are published, due tochanges in technology and advances in cognitive science.

• Researchers should consider collaborating on “model games.” Suchgames would enable controlled research studies in which investigators develop variations on the models and test them among differentgroups of learners to address a suite of related research questionsabout factors that may influence the effectiveness of games as learning tools. New model games would build iteratively on old models,based on this research.

Studies of the effectiveness of simulations and games for science learning have tended to focus on assessing conceptual understanding alone. The research has given little attention to the broader science learning goals advocated by science education experts. Research is needed to improve understanding of how simulations and games can best motivate learners, engage them in active investigations, and build understanding of science processes as well as concepts.

• Researchers should assess the potential of games and simulations toadvance a broad set of science learning goals, including motivation, conceptual understanding, science process skills, understanding of the nature of science, scientific discourse, and identificationwith science and science learning. Such research is needed to moreclearly illuminate the full range of science competencies that can besupported with simulations and games.

This report has shown that simulations and games have potential to address critical weaknesses in current science education by meeting the individual learning needs of both low-achieving and advanced science students, embedding science learning in the context of engaging real-world problems, and improving access to high-quality science learning experiences in formal and informal settings. An important first step toward reaching this potential is to increase basic understanding of the processes of learning when individuals interact with simulations and games.

• Research should examine the mediating processes within the individual that influence science learning with simulations and games.This research would aim to illuminate what happens within theindividual—both emotionally and cognitively—that leads to learning and what design features appear to activate these responses. Forexample, a game may arouse an emotional response and/or encourage the learner to set goals. Over time, such studies might begin toidentify the ways in which different design features activate sharedemotional and cognitive responses that support science learningacross individuals.

• Research on games should seek to develop empirical links betweendifferent types of motivators and different learning outcomes. Forexample, extrinsic motivators, such as points or opportunities toadvance to a higher level of game play, may encourage learnersto repeat and remember important science or mathematics facts, whileintrinsic motivators, such as satisfying one’s own curiosity or interest,may motivate deeper conceptual understanding and development ofscience process skills. Social motivators, such as the desire to participate or to establish an identity in a group of game players, might beparticularly effective in encouraging the development of scientificdiscourse and identification with science and science learning.

• Research should examine the role of metacognition and awareness ofoneself as a learner when an individual interacts with a simulationor game. Prior research on science learning suggests that makinglearning goals explicit and supporting learners in metacognition—reflecting on their own learning—enhance learning. In contrast,simulations and games can be designed to support “accidental” learning through playful engagement. Research is needed to determinewhether, and to what extent, science learning may take place evenif the learner is not aware that he or she is engaged in learning.

• Studies are needed to explore which individuals and groups preferwhich types of simulations and games for science learning, as wellas the durability of such preferences. They should consider howindividual preferences are related to individual personality traits,broader group characteristics, the nature of the learning experienceitself, learning processes, and learning goals. These studies shouldalso consider how context and experience can broaden or changeindividual and group preferences.

• Researchers should establish stronger theoretical underpinnings forthe use of simulations and games by connecting research on simulations and games to the relevant theory and research on learning

more generally, drawing on social and cognitive psychology, human-computer interactions, anthropology, and other fields that studylearning.

Although simulations and games provide contexts that can motivate and support learning, research on games has shown that learners may focus on the context or narrative to an extent that slows development of a deeper understanding of science concepts. Research is needed to explore this tension and illuminate how best to create virtual contexts that both motivate learners and support durable, transferable learning.

• Studies should examine how learning contexts created in simulationsand games may advance or hinder attainment of different sciencelearning goals. For example, engaging students in the context of avirtual investigation of a real-life problem may simultaneously advance multiple learning goals (e.g., conceptual learning and scienceprocess skills), or it may advance one or more goals while having noeffect on slowing attainment of others.

• Future studies should examine transfer of learning from the simulation or game learning environment to other contexts. These studiesshould examine how transfer occurs (including the features ofsimulations and games that support transfer), the extent of transfer,and whether including data drawn directly from the real world insimulations and games influences students’ understanding of scienceprocesses and/or motivates them to make real-world decisions basedon evidence.

• Research is needed to examine the durability of science learning thatis advanced through interaction with simulations and games. Forexample, some individuals develop feelings of identity with scienceand science learning through extended interactions with games.Investigators should track such individuals over several years toassess the extent to which this identification with science translatesinto sustained science achievement. In addition, they should conductretrospective studies to assess the extent to which identity with science developed through gaming may encourage entry into sciencecareers.

Overcoming current barriers to the use of simulations and games to help all students learn science requires research and development in a number

of areas. This section of the research agenda focuses on research related to learning; in later sections, the committee recommends research to understand and mitigate constraints to wider use of simulations and games.

• Future research should investigate how simulations and games cansupport diverse learners in science and mitigate particular individualor group learning difficulties, such as lower science achievementlevels, limited English proficiency, lower general cognitive ability,learning disabilities, or attention deficit hyperactivity disorder.

• Research should examine whether, and to what extent, diverselearners develop intuitive understandings of science processes andscientific modeling through play in the model-based virtual worldsof recreational games and how games designed for science learningcan build on these intuitive understandings to develop knowledge ofscience processes and the nature of science.

Simulations and games have potential to enhance science learning in formal contexts, such as science classrooms or online science courses, and in informal contexts, such as homes, after-school clubs, libraries, and recreation or science centers. Research to date has shown that the context significantly shapes how learners interact with a simulation or game and the extent to which this interaction supports science learning. Further research is needed to more fully understand how different contexts affect learning with simulations and games and to investigate how the design of learning environments might impact learning. To supplement the research recommended above, which would use model games to assess the influence of different contexts, researchers should

• Investigate how best to integrate games into formal learning contexts(K-12 and higher education) and informal learning contexts (e.g.,home, science museum, after-school club) to enhance learning. Thisshould include studies of how internal scaffolds in the simulation orgame and external scaffolds provided by a teacher, mentor, peers,or other instructional resources (either in person or via various onlinemechanisms) support science learning in different contexts.

• Examine current policy and practice barriers that slow the adoptionand use of high-quality simulations and games for science learning inK-12 and higher education. This research should include examination of such barriers as the need for teacher and faculty professionaldevelopment and the limited availability and quality of assessments;technological barriers, and barriers to research in real-world settings.

Studies of barriers in K-12 education should examine the role of current state science standards and accountability systems as barriersto increased use of simulations and games.

• Examine social and cultural factors in both formal and informallearning contexts that influence how widely simulations and gamesare used for science learning. Investigators should examine howchildren and adolescents, parents, caregivers, informal educators,teachers, school administrators, and education officials perceivethe educational and entertainment value of games and how theseperceptions may enhance or limit wider use of games designed forscience learning. The findings of this research should be used todevelop targeted solutions that should then be tested for effectivenessin intervention research.

• Examine the potential of different types of simulations and games,as well as different types of delivery platforms, to bridge informaland formal science learning. This should include research on thepotential of “lightweight” games that can be easily accessed on the webusing cell phones and other mobile devices to support learning acrossboundaries of time and space.

• Study the potential of structured informal learning environments,such as after-school clubs and online learning communities, aspromising contexts for science learning with simulations and games.Such studies should examine how learning in these environmentsmay transfer to or support further science learning in the classroomand at home.

• Study how engaging learners in implementing or modifying existingscience learning games or designing new science learning gamesmay advance one or more science learning goals.

Research on how to effectively assess student learning with simulations and games and use that information to impact the learning process is still in its infancy, although initial work seems promising. Achieving the potential of simulations and games for assessment and learning will require research and development in all areas of assessment: development, implementation, and evaluation. In particular, research is needed on:

• Applications of the evidence-centered design approach to the development of assessments of learning through simulations and games.Developers and testing experts should collaborate to clearly identifydesired learning goals and the kinds of evidence needed to showlearner progress toward these goals; they should use these specifica-

tions to design tasks and test items in ways that will provide the neededevidence. Modeling of the motivation and thinking of the learner willneed to evolve simultaneously with the “physical” modeling of thegame or simulation.

• The development and use of flexible statistical models and machinelearning to make meaning from the large amounts of data providedby simulations and games. These measurement methods are wellsuited to application in simulations and games, because they canhandle uncertainty about the current state of the learner, provideimmediate feedback during tasks, and model complex patterns ofstudent behavior and multiple forms of evidence. Continued researchon these methods will help to improve assessment in simulations andgames.

Assessment tasks seamlessly embedded into game play and linked to instructional supports have great potential to support individualized science learning. Simulations and games can be designed to rapidly interpret learner performance on these tasks, using the information to provide the learner (and teacher) with feedback, coaching, or new information or learning challenges, based on the student’s unique capabilities and learning needs. These promising developments, if supported by further research, could lead to radical improvements in self-directed science learning and the authentic assessment of science learning.

• Researchers should continue to advance the design and use of techniques that (1) rapidly measure and adapt to students’ progressin a specific learning progression, (2) dynamically respond to anindividual student’s performance, and (3) allow for the summativeevaluation of how well students are learning.

The committee identified two possible models for reaching scale in the use of simulations and games for science learning in formal education: (1) a traditional top-down market model, in which games or simulations are sold or distributed to universities, schools, and school districts, and (2) a market model in which widespread use of simulations and games for informal science learning by parents, students, and individuals could dramatically change how science is learned and taught in schools and colleges. Neither model can become reality without research to more clearly illuminate the current barriers to implementation and to identify approaches to overcoming these barriers. For example, there is not yet a coherent market for either games or simulations in schools that is analogous to the textbook market,

and the bewildering variety of games and simulations for science learning available for free or for purchase can leave potential customers confused. The committee recommends the following:

• Research to better understand key factors that will enable both theeducation marketplace and the informal learning marketplace toembrace games and simulations for science learning. The goals ofthis research should be to increase understanding of key design features that enhance the appeal and uptake of games and simulationsand market forces that affect adoption across formal and informallearning contexts.

• Research and development partnerships should be established to investigate alternative mechanisms for supporting large-scale collaborative innovation in science education based on the use of simulationsand games and to support ongoing improvement in simulations andgames.

• Research on the feasibility of systems for informing users or consumersabout the quality and educational effectiveness of simulations andgames designed for science learning, such as expert rating systems.This research should explore the potential of such systems to serve ascatalysts for distribution of high-quality simulations and games.

To carry out all elements of this research agenda, the committee recommends creating research and development partnerships:

• Academic researchers, developers and entrepreneurs from the gamingindustry, and education practitioners and policy makers should formresearch and development partnerships to facilitate rich intellectualcollaboration. These partnerships, which may be large or small,should coordinate and share information internally and with otherpartnerships and should

  • share resources and tools, thereby reducing costs and allowingreusability;

  • provide researchers with shared points of access to students andtheir educational records and to informal learners, at the sametime conducting research that assists formal and informal learning institutions;

  • explore alternative approaches to—and economic models for—extending the life cycle of simulations and games with ongoingupdating and maintenance; and

• investigate how to optimize educational contexts for simulationsand games—including alternative technologies and platforms,teacher preparation and professional development, and curricularsupports—for different populations of K-12 and adult learners.

• Government agencies and foundations may consider the potentialbenefits of providing sustained support for such partnerships.

• Government agencies and foundations may consider the potentialbenefits of funding research and development of new models fordelivering learning opportunities through simulations and gamesthat can be self-sustaining and reach a broad audience.

• Researchers in the software and gaming industries, governmentagencies, and academic institutions should continue their researchand development of new, open-source authoring tools to facilitatedevelopment of games and simulations.

The research agenda outlined in this chapter is meant to provide guidance to active and prospective researchers, simulation and game developers, commercial publishers, and funders. However, games and simulations designed for science learning are played and used by a wide variety of individuals in rapidly changing markets. In the future, this research agenda may change with advances in technology, shifts in consumer preferences, and changes in the education environment. The committee expects that, if implemented, the research agenda will have to adapt and evolve in tandem with the evolution of the field of educational simulations and games.