In the fall of 2015, I began the Martian Music Invasion (MMI) project with Jennifer Han under advisement of Dr. Amy Ogan. Through this project, we aim to discover the effects of curiosity on motivation and learning in educational games. Specifically, we want to compare motivation and learning when students’ curiosity about a game is integrated with educational content versus when curiosity is not integrated.
We designed, programmed, and user-tested Martian Music Invasion, a game about rescuing and rebuilding a song, which was scattered into the sky by a group of evil Martians, so that Earthlings can continue to enjoy beautiful music. Players direct a hero to pick up her minions, which each possess musical energy for a particular tone A through G, and place them in the correct position on a musical staff to match the song. Our target demographic is 8-10 year olds who haven’t learned how to read music.
Curious about which one you should play? Start with the integrated curiosity version. I think it’s more fun.
Acknowledgements ▲ Hide
We could not have done any of this without Dr. Amy Ogan. She teaches the class on educational game design and is an author of the paper introducing the EDGE framework for designing educational games. She guided Jennifer and me through every step of the process, from literature review to brainstorming to paper prototyping to getting feedback to playtesting to programming to more playtesting. Going in, we both knew very little about most of these steps. Amy always knew just what to tell, ask, or show us to make sure we made it through each step in the few weeks we had for it. Jennifer and I are both very grateful to have the opportunity to work with and learn from Professor Ogan.
Thanks to Erik Harpstead, a PhD student working on data science methods that can be used to analyze educational games. Erik advised us on a number of aspects of our game, including the tradeoffs of different methods of character movement, transfer of learning from within the game to outside the game, and the different available methods of collecting results about players’ learning, motivation, and curiosity.
Thanks to Dr. Jessica Hammer, an assistant professor at the HCII and ETC who studies games. She gave us important feedback about our experimental design and conceptualization of curiosity, explained in depth in the Design section of this article.
Thanks to Alexandra To, a PhD student in the HCII studying curiosity. Alexandra helped us with our experimental design. She helped us to understand how to quantitatively measure curiosity and to design control and experimental test groups.
Thanks to Gail Kusbit, the HCII research manager. Gail helped us find 8-10 year olds who haven’t learned to read music to playtest our game, gave us pointers on communicating with parents, and as a piano teacher also gave us insight into music education!
Background ▲ Hide
Many educational games don’t integrate educational content with gameplay at all. An extreme example of this is Note Name Invaders, which is a game like Space Invaders where the player is quizzed on their knowledge of music note names. The game is interrupted every few seconds with a question about a note name, and the player needs to answer it correctly to continue playing. This interruption breaks the flow of the game, reducing the player’s engagement and motivation to keep playing. Even worse, players see gameplay as an external reward for answering questions correctly. Extrinsic rewards have been shown to undermine motivation to learn, so while this game may improve scores on a quiz in the short term, it will contribute to students’ loss of interest in reading music in the long term. However, there are advantages to this type of game. The game designer can come up with a compelling design without restrictions imposed by having to include educational content and without knowledge of what content that will eventually be included in the quiz. Additionally, the game can be reused over many subjects without incurring any additional game development costs; only the quiz questions need to be changed.
In the scope of this project, we are interested in games that foster strong player engagement and do not offer extrinsic rewards to learning. We are interested in games that use intrinsic motivation to play directly as intrinsic motivation to learn. Gameplay must be integrated with learning. The drawbacks of this type of game are that the game must be designed with a specific set of educational goals in mind and the game cannot be reused with different goals.
Some educational games have been extremely successful, such as Oregon Trail, Civilization, and Sim City. TeachThought explains in-depth what made these games, and other, successful. A common theme for most of these games is that the educational content is intrinsically linked with either the fantasy context, game mechanics, or goals of the game. What you learn from it is the same thing that makes it fun. In Oregon Trail, the fantasy context, a historical journey, is the educational content. In Lemonade Stand, the game mechanic, running a business, is the educational content.
There has been some fascinating research on integrating educational content with game mechanics, such as Habgood & Ainsworth’s Zombie Division study. In their study, they developed a game about destroying skeletons using weapons. In one version of the game, students alternated between playing the game and practicing math problems. In the other version, the skeletons and weapons were assigned numbers and the player could destroy a skeleton only if the weapon’s number divides evenly into the skeleton’s number. In the second version, when the game had integrated game mechanics and educational content, students were more engaged (played for longer, given the choice) and learned more (scored higher on a post-test.)
In our research, we found four main components of games that make them intrinsically motivating (i.e., fun:) game mechanics, fantasy context, goals, and curiosity. While we found some research on curiosity in games and plenty of research on curiosity in education, we found very little on curiosity in educational games and none on integration of curiosity in educational games. We took the Habgood & Ainsworth study as inspiration, which examined integration of game mechanics in educational games, and chose to do something similar with curiosity. We want to examine the differences in motivation and learning between two versions of a game: one where curiosity is integrated with educational content and one where it is not.
What do I mean by curiosity? There are lots of similar but different definitions of curiosity in circulation. Broadly, curiosity can be thought of as the drive, motivation, or action to seek or resolve sensation or knowledge that challenges or expands upon one’s experiences. The most important thing to notice here is that for curiosity to exist, one must know that one’s experiences, whether sensory or cognitive, are incomplete or inconsistent. One cannot have the desire to challenge or expand something which one believes is consistent and complete. Therefore, to inspire curiosity in somebody, you must show them that their knowledge or sensory experiences are incomplete or inconsistent, that their is a gap in their current information.
With that idea of curiosity in mind, we can think about how to integrate curiosity about a game with curiosity about educational content. There is a domain of ideas and symbolic relationships contained within the world of every game. This domain encompasses the fantasy context and game mechanics. Games that explain everything to you at the very beginning do not foster much curiosity. Games that reveal more about their lore as you find new regions of the world or drop hints that your knowledge of the available tools and powerups is incomplete do foster curiosity. We can foster curiosity about educational content by showing students that there is more that they have yet to learn. If the concepts that they have yet to learn are simultaneously elements of the gameplay or fantasy context that they have yet to discover, then their curiosity is integrated. Otherwise, even if students are curious both about the learning content and the game, if these curiosities are not the same then the curiosity is not integrated.
Design ▲ Hide
Our design process began with the goal of creating two versions of a game with different levels of integration between curiosity about the game and curiosity about the learning content. We quickly decided on reading music as the educational content. We answered some important questions:
- Will this game introduce new content, or simply help students practice what they already know?
- Since we want players to be curious about content they haven’t learned yet, the game will introduce new content.
- What exactly are the learning outcomes for this game?
- Players should be able to name a note on a treble or bass clef staff based on its position.
- Are there any areas of potential transfer – things that players may learn from the game but are not necessary to complete the game?
- Players may learn about the mapping from note name/position to pitch, note lengths, rests, and the fact that the bass clef has lower pitches than the treble clef.
Next, we moved into an brainstorming phase where we considered many combinations of what genre of game we wanted, what the game mechanics should be, what fantasy context to use, what type of curiosity to use, and how to design the game around integrating curiosity.
Click on an image to learn about its importance in the brainstorming process.
After brainstorming, we came up with this initial concept for the game:
We talked with Dr. Jessica Hammer about this idea and how we hope it will foster different types of curiosity and she pointed out some significant issues. There are far too many differences between the integrated and non-integrated level selection screens. Differences in results between experimental test groups could come from any number of confounding factors, and not necessarily the differently natured curiosities. The integrated version has a clearly linear unlocking progression, whereas the non-integrated version does not. In the integrated version, the player unlocks pieces of the song, which we want, but in the non-integrated they are not shown pieces of the song at all. This extra exposure to written music and the tighter connection between the level play and level selection could lead to greater motivation and learning, regardless of the effects of curiosity.
Dr. Hammer also prompted us to refine our conceptualization of curiosity and to be able to better articulate exactly what we want players to be curious about and why. We want them to be curious about the parts of the game that are locked, including any not-yet-introduced concepts in either the sheet music or the comic. Did we want to measure players’ curiosity levels in general, about music, or about the parts of our game that are locked? We decided to measure curiosity about the game, since we want to know whether an increase in curriculum-integrated curiosity about the game yields a greater increase in learning and motivation than does an increase in non-integrated curiosity.
We refined our design to manipulate as little as possible: the only thing manipulated between the two versions is whether the part being unlocked is a song or a comic strip. Both versions include partial exposure, an information gap, a comic, an audible song, and sheet music for the song.
We then built paper prototypes. The purpose of the prototypes is twofold: first, building paper prototypes forced us to make detail-level design decisions, some of which turned out to be more important than we expected. Should the hero be moved manually (i.e., arrow keys) or by directively (i.e., clicking where she should end up)? Should you be able to turn in multiple notes at a time? How many lives should you start with? When should we introduce new concepts? How? What should be the story in the comic strip below the sheet music? What should go on cutscenes? Why minions? Did the Martians take all the music from Earth or just some of it?
When we first came up with the concept and explained it among ourselves and to others, we brushed over many of these details. Explicitly drawing out everything that players will see before, during, and after gameplay forced us to think about all of these decisions. Doing this before we started coding made it easier to work directly with the design. By drawing the design, we were closer to its feel than we would be by coding.
The second purpose of building the prototypes is to test the idea with kids before setting it into a less malleable form (code.) Here are the paper prototypes we created:
After testing these paper prototypes with kids, we built the electronic version of the game using Unity and ran some more playtests using this version.
Future Work ▲ Hide
Jennifer and I will continue working on this project with Dr. Ogan in the spring of 2016. We will continue to iterate on the design of the game and add more “polish” features to make it more appealing.
As we polish the game, we will also be looking for a school with a class of 2nd-4th graders that we can run a larger scale experiment with. We will test both versions of the game, integrated curiosity and non-integrated, and measure the differences in learning outcomes and free-choice motivation to determine what difference, if any, is effected by altering the nature of curiosity-arousing elements in an educational game.