JISC

A growing body of research indicates that playing computer games can lead to serious learning. We propose however that the learning can be deeper and more richly interconnected if game playing is combined with game making. Professional game developers must bring together expertise on game design with knowledge about the game's subject domain, which can require expertise in mathematics, science, and computation. Furthermore, game development frequently requires a good deal of expertise in art, music, drama, film, and narrative. We believe that experiencing even a tiny part of this game making process can give learners a deep and broad learning experience. However, acquiring the skills to make computer games requires a major investment in time and effort. In an educational setting, this large investment in acquiring the prerequisite programming skills can sometimes be justified by the potential payoff in general problem-solving and model building skills, but the need to acquire sophisticated programming skills is still a daunting barrier to the adoption of game making in a general education setting. 

Here, however, we propose to build upon some novel research in how games can be made without a priori programming skills. Rather than building games and simulations by programming from scratch, the student can make use of specially crafted software construction kits. These kits are designed to support a genre or class of games and simulations. One construction kit may support the creation of space games for exploring Newtonian Mechanics and Orbital Mechanics, while another may support the exploration of flocking and herd behaviour as studied in a Zoology course. Even learners who lack computer-programming skills can use these construction kits.

A computer game can be analysed as consisting of a simulator together with a competitive or cooperative play component. We intend to support the construction of simulators that model some subject matter that is taught at the university level. Depending upon our choice of subject matter and the preferences of our faculty collaborators the game play component may be downplayed or even non-existent. We see the construction of simulators as the crucial component pedagogically. The addition of a game play component is not fundamental but may significantly increase the motivation and appeal of the learning designs.

One promising method to motivate, support, and structure game making is to present the students with a metagame within which games are built. A metagame is an instance of a design for learning – it consists of entities central to the IMS Learning Design specification: roles, activities, resources, and services. Metagames can be constructed for any subject matter for which simulations or games can be built. This covers much of science, sociology, history, economics, government, business, and perhaps more.

In a typical metagame, the learner takes on the role of a member of a game development team that includes programmers, scientists, historians, artists, game designers, and more. For instance in a prototype developed by Kahn for the BBC Digital Curriculum project these roles were implemented as simple software agents. The learning design underlying the metagame consists of three components:

1. assignment of tasks (typically involving game making or simulation creation) with supporting resources
2. interactions with team members in order to acquire additional necessary resources
3. interactions with team members in order to acquire in-depth and background knowledge

An important aspect of a metagame is the conditional behaviour of the simulated team members. These agents provide information and resources, taking into account the current state of the learner’s constructions and the history of previous interactions. Each simulated team member provides contextually appropriate support, context, and background information from a unique perspective: team members can be physicists, historians, computer scientists, game designers, or mathematicians. Furthermore, unlike a typical learning design player, a metagame can have the appeal and the look and feel of a video game.

Learning by constructing games and simulations is an instance of constructionist pedagogy. As learners construct models, simulations, and games, they are also constructing their understanding of the underlying subject mater. Software construction kits are well suited for layered learning, in which learners are the locus of control. Even quick and superficial game construction will typically result in a worthwhile learning experience. If, however, learners delve further to acquire a deeper understanding and control, the result will be a much richer learning experience. This learning is typically in a social context where the learners may be collaborating in real-time, either face-to-face or via a network, receiving support via asynchronous forums or Wikis or being guided by teachers.

The prototype developed for the BBC demonstrated the promise of this approach. It did not, however, address many issues of importance to the learning design community. Both the content and learning designs were embedded in the software; a better design would involve a standard form that non-experts could create, reuse, or edit. There was no provision for on-line collaboration, support, or sharing of knowledge, components, and advice. The prototype addressed the learning of the physics of motion at a level appropriate for 10- to 14-year-olds, but avoided topics more appropriate to higher education.

There is general agreement that IMS Learning Design needs to be extended to support game playing: for example, Olivier and Tattersall propose that “sophisticated simulation and multi-user game-playing systems” should be added to the Learning Design services specification.  The DfES e-learning strategy report states,

'borrowing ideas from the world of interactive games, we can motivate even reluctant learners to practice complex skills and achieve much more than they would through traditional means'.

It is the aim of this proposal to remedy these shortcomings. In doing so, we will learn where the current set of ideas underlying the IMS Learning Design specification can be fruitfully applied and where there is a need for further development. Game and simulation construction kits require learning design structures that are much richer than the linear sequences commonly used. We will pioneer the design of learning structures that rely heavily upon branching and concurrent activity.

Aims and Objectives

• Explore the promising area of simulations and games in the context of designs for learning.
• Explore learning designs in which learners collaboratively construct computational objects.
• Further develop the open source LAMS Activity Management engine to explore areas that have been highlighted by practitioners as important: conditional workflow and the integration of news tools or services.
• The dissemination and workshop project activities will enable practitioners from around the UK to engage with this area of research.
• The focus of prototyping tools that support design for learning at Oxford will enable a better understanding of the relevance of software in this domain at Higher Education level.

Project Methodology

The LAMS 2.0 release incorporates a new architecture (see figure 1) in which each activity/ service uses a well-defined interface to the LAMS authoring, administration, monitoring, and runtime engine. This means we can build a layer of services over the simulation construction tool we will build to allow teachers to define and coordinate activities that utilise this tool for learning.

We will pilot the development of designs for learning by constructing computer models and games in collaboration with Oxford University faculty. We will test our designs in the context of undergraduate learning at Oxford. This is in keeping with the university’s strategic plans, which aim to

'stimulate the creation of new patterns of teaching and assessment … which focus on learning outcomes' and to 'encourage changes in course structures that would allow increased student choice and mobility [and] innovation in programme design'.

We plan to explore new kinds of designs for learning by constructing computer models using LAMS (work package 2) and building new activity tools that support the construction of games and simulations by learners (work package 3). We will explore questions of generality and usability by studying the university curriculum and meeting Oxford faculty (work package 1) and by hosting two workshops (work package 4 and 6).

We plan to begin our exploration of possible subject domains by studying various Oxford University course descriptions, lecture notes, and textbooks. We will meet with faculty members to acquire subject expertise and collaboration. We will host two open workshops to present and discuss designs for learning by building computational models and games.

Simultaneously, we will investigate the use of LAMS as a tool for supporting learning designs based upon model construction. We expect to discover shortcomings that will inform the next phase of this research, and we will provide recommendations for enhancing the learning design frameworks and tools.

We will build a prototype simulation construction tool that will build upon the ideas underlying the space game construction kit built for the BBC, and the findings of the earlier work packages. This prototype will differ significantly from the BBC prototype however, particularly in its emphasis on reuse and building upon community standards. We plan to build this prototype as a LAMS activity tool where LAMS learning designs will provide the structure, guidance, and monitoring of the use of our new constructionist activity tool in an open and reusable manner.

As soon as the prototype becomes operational, we will begin user testing. This will serve several functions including providing guidance for the iterative design of the software and the learning designs.

Deliverables

• Source code and compiled distribution of simulation construction toolkit
• Documentation for technical and pedagogical audiences made available through an open access wiki
• Two open invitation workshops
• At least 10 designs for learning by constructing model expressed as LAMS sequences

Stakeholders

• Instructors in fields where there is active multi-agent simulations research, e.g., zoology, sociology, economics, computer science, and earth sciences
• Both undergraduate and graduate students in those fields
• Researchers in the field of computer modelling tools
• The LAMS developers and community
• The JISC community
• Learning technologists

Other information

The project website is a Wiki that we using to document the evolving design, project meetings, scenarios, and much more.